In vitro testing

of inorganic phosphorus sources for

phosphorus availability in swine.

A thesis submitted in fulfillment of the requirements for

the degree of Master of Applied Science

John Cauduro

B App Sci (Applied Chemistry), AssDip (Applied Biology)

School of Applied Sciences

Science, Engineering and Technology Portfolio

RMIT University

March, 2009

ii

Acknowledgements

My sincere thanks is extended to my principal supervisor Dr Nichola Porter for her

guidance, assistance and encouragement throughout my study. I also wish to thank my

co-supervisor Dr Barry Meehan for all his support particularly at the start of the study,

and to the other students and staff at RMIT University.

I would like to acknowledge the staff at the Department of Primary Industries (DPI)

Werribee for their service and support in setting up and conducting the animal segment

of the project, as well as the use of their equipment. In particular, I would like to thank

Dr Ray King for being the Principal Investigator and Co-ordinator of in vivo studies and

for the overall care and responsibility for the project, including the provision of funding

obtained from APL (Australian Pork Limited).

I would also like to acknowledge DPI for their time and assistance especially with

technical support. Thanks also go to to Paul Eason, Matthew Borg and Peter Bisinella,

for their assistance, predominantly with the nurture of the animals, and providing the

relevant samples. To Matt Kitching for training on the MIR and the GRAMS-AI PLS

software. To Tadeusz Golebiowski from DPI Horsham for running the NIR and assisting

with some of the software analysis. To Dr Iko Burger from CSIRO Clayton for the use of

the SS-MSA-NMR.

Many thanks for the donations of samples and feed materials from APL, DPI Werribee,

QAF and the other suppliers of the feed ingredients.

iii

Finally to my family and lovely wife Melanie and her family, for all of their support and

help.

Declaration:

I certify that except where due acknowledgement has been made, the work is that of the

author alone; the work has not been submitted previously, in whole or in part, to qualify

for any other academic award; the content of the thesis is the result of work which has

been carried out since the official commencement date of the approved research

program; any editorial work, paid or unpaid, carried out by a third party is acknowledged;

and, ethics procedures and guidelines have been followed.

Signed

John Cauduro

Dated.

iv

Table of Contents

Acknowledgements ………………..……………..………………………………………………….….ii

Statement of authenticity ………………………………………………………………………..……...iii

Table of Contents …………………………………………….……………………………..…………..iv

Table of Figures………………………………………………………………………….……………....vi

Table of Tables ……………………………………………………………………………………….…vii

Table of Equations …………………………………….……………………………………….…........viii

Appendices ………………………………………………………………..……………………..…. ….viii

Glossary and Abbreviations used in this Thesis ……..…………………………………………...…..ix

Abstract ……………………………….……………………………………………………….…………..x

1 Introduction ……………….…….……………………………………………...…….……….….1

1.1 Phosphorus…………..…………………………………………….……………..……..2

1.1.1 Phosphorus pollution…………………………………….….………………..2

1.1.2 The Pig industry…………………………………………...……………..……3

1.1.3 The Project……………………………………………….…………..………..4

1.2 Pig nutrition ………………………………………………………..……..……………..5

1.2.1 The need for phosphorus…………………………………..………..……….5

1.2.2 Amount of phosphorus needed ……………………….…….………………6

1.2.3 Reduction of phosphorus excretion………………………………………....7

1.3 Feeds…………..……………………………………………………..………………...10

1.3.1 Inorganic Phosphorus sources……………………………………………..10

1.3.2 Organic Phosphorus sources ……………………………………………...12

1.4 Analysis of digestibility……………………………………………………………..…14

1.4.1 In vivo Digestion…………………………………………………………..…14

1.4.2 In vitro Digestion………………………………………………………….….16

1.4.3 Spectroscopic techniques ………………………………………………….18

v

1.5 Research outline …………………………………..…………………………...……25

1.5.1 Rationale for research ……………………….…………………….…...…25

1.5.2 Objectives ………………………………………….…………………….…27

1.5.3 The project description…………………………………………………….28

1.5.4 In vitro methods chosen …………………………………….…………….29

2 Materials and Methods …………….…………………….……….…….………………….…32

2.1 Collection of the feed additives ……….………..…….…….…………………..….33

2.2 Testing of the feed ………………………………..….……….………………….….34

2.3 Preparation of the feed ……………………………………....………………….….34

2.4 The animal testing……………………………………………….………..…..……..38

2.5 The selection of the pigs…………………………………..…..….…………………39

2.6 Feeding of the animals..…………………………………..………………..………..40

2.7 Collection of the faeces and urine..…….……………….……..…..……….………40

2.8 Testing Sites ……………………………………………………..………..………….42

2.9 Calculation of P – Digestibility …………………………………..…………..……...42

2.10 Digestibility by difference …………………………………………..………………..45

2.11 Invitro testing………………………………..………………….……..………………47

2.12 Methods for testing …………………..………………………………..…………..…47

2.13 Wet Chemistry Methods …………………..………………….………..…….……...47

2.13.1 Chemicals and reagents………………………………………..…………..48

2.13.2 Procedures………………………………………….……………..…………53

2.14 Non – Destructive Methods ……….…………………………………..…..………...57

3 Results & Discussion .……………………………………………..……………………...…...61

3.1 Outline ………………………………………………………………………………….62

3.2 The diet and its ingredients…….………………………..……………………….…..63

3.2.1 Ingredients ……………………………………………………………………63

3.2.1 Dietary mixtures ……………………………………………………………..67

vi

3.3 The analysis of urine and faeces…………………………..……………….………..70

3.3.1 The analysis of P in Urine and Water ……………………………………..70

3.3.2 The analysis of Faeces …………….. ……………………………………..72

3.3.3 The analysis of Ca in Urine…………. ……………………………………..75

3.4 Pig body weight……………..…….…………………………..…………..….……….76

3.5 Calculation of P-Digestibility…..……………………………..………….….…...…..79

3.6 The in vitro wet chemistry tests.………………………………..…………………...86

3.7 The infrared analysis..……..…….………………………………..……….….……..94

3.8 Solid State 31P Magic Angle Spinning NMR……………………..………….……103

4 Conclusions ……………….…………………………….……….…………………..………108

4.1 Conclusions ………………………………………………………………………...109

4.2 Future work …………………………………………………………………………113

4.3 Implications ……………………………………………………..…………..………115

4.4 Recommendations …………………….………………………….…………..……115

References ..………………………………………………..………………………….…………….…117

Appendixes…………….…………………..………….……..………………………….…………..….130

Table of Figures

Figure 1. Brands of feed additives used in the trials ……………………………….……....33

Figure 2. Feed hoppers containing the feed mixes for this project ……………….………38

Figure 3. Mixing bowl used for homogenizing the faeces ………………………….………41

Figure 4. Solid State NMR pulse sequence…………………………………..……….……..60

Figure 5. Total P concentration (mg/L) in pig urine for each diet type ……….…………...70

Figure 6. Total Ca concentration (% w/w) in pig faeces for each diet type …………….…72

Figure 7. Total P concentration in (% w/w) in pig faeces for each diet type …….………..73

Figure 8. Total Ca concentration (mg/L) in pig urine for each diet type ……….……..…...76

vii

Figure 9. Pig Weight Trends ..………………………………………………………..……......77

Figure 10. Comparison of the in vitro test for P digestibility ………………………..…….….87

Figure 11. Comparison of extractant times (1/2 Hr and 16Hrs) and pH …………..…….….88

Figure 12. Comparison of sample mass to extractant volume ratio………………..…….….90

Figure 13. Comparison of water and calcium chloride solution……………………..…….….91

Figure 14. MIR spectrum or rediphos DCP and honghe DCP ……………………..……......95

Figure 15. SS 31P MAS NMR of rediphos DCP ……………………………………..……….105

Table of Tables

Table 1. Weight of Individual feed ingredients (kg) ..…………………………….……….....36

Table 2. Balances and their accuracy…...………………………………………..……………37

Table 3. Randomized allocation of diets ...…………………………………………………....39

Table 4. Test Methods weights, volumes and extraction times………………….………….56

Table 5. Calcium and Phosphorus Concentrations in the feed ingredients .…………..…..64

Table 6. Dry Matter (DM) content, dried at 100 oC for the P-sources ..…….………..….....66

Table 7. Calcium and Phosphorus results of each diet type…….. .…..……………….……67

Table 8. Dry Matter (DM) content, dried at 105 oC for the Diets ..………….……………….69

Table 9. Calcium : Phosphorus Ratio in pig faeces for each diet type …………………......74

Table 10. Pig numbering allocated against the pig identification……..……………………....77

Table 11. P - Digestibility results for each pig …………………………..…………..…............79

Table 12. Average basal digestible P result for each run, and t-test for run 2..……….……..81

Table 13. Dietary P Digestibility, mean and standard deviation ………...………………….....82

Table 14. P Digestibility mean and standard deviation ………………….……………............83

Table 15. Correlation results between in vitro and in vivo P digestibility ……….…………….92

Table 16. Variation in in vivo P digestibility measured by MIR ………………………..............98

Table 17. Total P and Total Ca measured by MIR in all phosphorus samples ……............101

Table 18. Peaks and sidebands for 16 samples by SS 31P MAS NMR …………..………....104

Table 19. Suggested P digestibility (%) of a generic P source ……………………..………...116

viii

Table of Equations

Equation 1. Total Nutrient Intake (I) ...……….……………………………………………………43

Equation 2. Total Nutrient Output (O) ...…….…………………………………………………….44

Equation 3. Digestibility in the test diet (Digtest) ...……………………………………..…………45

Equation 4. Digestibility of the Nutrient in the Additive (Digadd) ...…………………….………...46

Appendices

Appendix A. Percentage of Ca, K, Mg, Na, P and S in the ingredients ………………………..130

Appendix B. Example of printout and calculation from ‘FEEDMANIA DIET’ …………………134

Appendix C. Weights of individual feed ingredients determined by ‘Feedmania’…………….135

Appendix D. Mineral results in the pig diets in mg/kg with mean and SD …………..………..136

Appendix E. Weights of Pigs in kg on the date shown in 2004 ……………………….…........137

Appendix F. Comparison of the in vitro test for P digestibility ………………………………....138

Appendix G. MIR spectrum………………………………………...……………………….………140

Appendix H. MIR Factor estimate curves for Digestibility, Total P and Ca …………..............143

Appendix I. SS 31P MAS NMR spectrum…………………………………….…………..……….146

ix

Glossary and Abbreviations used in this Thesis:

A: Amount of nutrient within the mix

Apparent the amount of phosphorus ingested minus the amount lost though faeces and

Digestibility: endogenously.

AOAC: Association of Official Analytical Chemists

Basal diet: Base feed, the diet without any additional nutrient additives

CSIRO: Commonwealth Scientific and Industrial Research Organisation

DCP: dicalcium phosphate or dicalcium phosphate dihydrate

DM: Dry Matter

DPI: Department of Primary Industries

F: Amount of faeces collected

FO: Amount of feed offered

Grower Pigs: Pigs which have been weaned and are in the process of becoming adults

I: Total nutrient intake, or the percentage of a nutrient consumed / ingested

MBM: meat and bone meal

MCP: mono calcium phosphate

MDCP: mono/dicalcium phosphate mixture

MIR: mid infra-red spectroscopy

MM: meat meals

N: Amount of nutrient NIR: near infra-red spectroscopy

NMR: nuclear magnetic resonance spectroscopy

O: Total nutrient output, or the amount of nutrient found in the faeces

P: Phosphorus

PRTC: Pig Research Training Centre

R: amount of feed refused

S: any feed spilt

SEM: Standard Error of the Mean

SS-MSA-NMR: Solid State-Magic Spin Angle-NMR

x

TCP: Tri calcium phosphate or rock phosphate

Abstract

This research project compares different chemical and spectroscopic techniques aimed

at finding a quick and cheap replacement for the measurement of digestibility of

phosphorus (P) in different inorganic feed additives in animal feed for pigs. This will

allow a reduction in the amount of feed additive used and in turn reduce environmental

problems caused by P in faecal pollution. This research has also yielded a comparison

of the digestibility of different feed additives.

P digestibility was determined from in vivo studies of pigs. The animal feed in the in vivo

studies contained P levels below the nutrient requirements. The basal diet was a corn

soybean meal base with the addition of amino acids, vitamins and minerals and the

addition of limestone for calcium (Ca) supplementation to approximately 0.8 % in all

diets. Assessment was performed on 6 different inorganic P sources, rock phosphate

(tricalcium phosphate (TCP)), meat and bone meal (MBM), mono/dicalcium phosphate

(MDCP) and three different dicalcium phosphates (DCPs). Analysis of the sources

indicated the MBM was at only 60 % of the expected amounts quoted for both total Ca

and P. Eight pigs where selected and placed into separate pens. Two were given basal

diets and the other 6 diets were randomly selected and supplied with the different

inorganic P sources. This was repeated over four consecutive 14 day trials (with the final

five day period of each trial conducted in metabolic crates for the collection of faeces). P

digestibility was calculated by difference (input minus output). Urine analysis confirmed

all pigs were deficient in P, ensuring maximum P uptake. The apparent P digestibility of

xi

the sources were: rock phosphate at 46 %, the MBM at 85 %, MDCP at 71 % and three

DCPs ranged from 49 % to 73 %. This substantiated that the apparent P digestibility in

the major inorganic sources of P is significantly less than 100%.

The in vitro or chemical methods of assessing phosphorus availability in animal feed

included the commonly used feed extraction methods of water solubility and 2 % citric

acid. These two methods showed significant differences between each other. Other

chemical methods used included calcium chloride, ammonium acetate, sodium

bicarbonate extractions, and a double extraction using hydrochloric acid (HCl) followed

by sodium bicarbonate. The majority of the chemical methods were not discriminative

enough. This was expressed by the very low and non-significant correlation coefficients

between the in vitro methods and in vivo P digestibility of the six phosphate ingredients

used. The citric acid method and the HCl / bicarbonate method using 0.25 g of sample

gave the highest correlation coefficients of 0.53 and 0.54 respectively, which were well

below the significance level. Citric acid dissolved over 95% of the DCP and MDCP,

whereas water dissolved less than 10 % for all the phosphates tested except the MDCP.

This could allow for differentiation between MDCP and the other calcium phosphates.

Infra-red spectroscopy is now commonly used in feed production for many other

nutritional tests. NIR, although being able to obtain an R2 above 0.999 for correlation

curves and factor prediction curves, could not obtain a self prediction of the calcium

phosphates due to the large Mahalanobis Distance. To make use of this method more

samples would be required for the calibration curve, and this research was limited to 6

samples. P digestibility predicted by MIR showed close agreement with the in vivo P

xii

digestibility. Again due to the small number of ingredients tested in the pig trial, the

prediction of digestibility using MIR could only be compared to it self. Hence MIR can

only be used as an estimate until more data can be obtained to confirm the predictions.

The correlation between in vitro and in vivo results for MIR was extremely promising (P

< 0.05) indicating that MIR was not significantly different from the in vivo P digestibility.

MIR predictions were also used to estimate total P and total Ca in the same way as was

done for the apparent digestibility. The results for total Ca and P were obtained

chemically. Additional samples obtained at a later stage were tested by MIR and for

total Ca and P. One of the additional samples, a mono ammonium phosphate clearly

showed errors in the MIR estimation of total P and Ca values. The remaining estimates

for total P were within 3 % (14 % variation) of the actual total result and the total Ca

results were within 5 % (32 % variation) of the actual total result.

The P31 SS-MAS-NMR indicated one of the DCPs was made up of 3 or more P

compounds by displaying 3 major peaks. All the P peaks in the faeces had different

positions to the P peaks in the ingredients, indicating some sort of change in the P form.

Overall the chemical methods were unable to predict P digestibility, and while the

spectroscopic techniques showed promise, they still require more work to examine many

more feed additives. It should be noted that the feed additives should be tested for total

P and Ca as the actual quantities may be different from the stated analysis.

1

Chapter 1

INTRODUCTION

2

1.1 Phosphorus

1.1.1 Phosphorus pollution

The process of eutrophication is a natural occurrence caused by increased nutrient

loading (in particularly phosphorus (P)) in a body of water producing excessive plant

growth. This is often followed by decomposition and subsequent depletion of oxygen

in the water body leading to fish kills. Inputs of P from animal manure to farm land

has accelerated the process of eutrophication, causing outbreaks of algae

(phytoplankton) including the toxic blue green algae varieties such as Anabaena,

Microcystis, Nodularia, Oscillatoria and Aphanizomenon. These algae produce

toxins (hepatotoxins (mainly microcystins), neurotoxins and endotoxins) which are

harmful to both animals and humans, causing large numbers of fish to die (fish kills)

and polluting water supplies (Kerr and Rayner, 1995).

Australia is particularly at risk due to its low rainfall and thus limited ability of its

waterways to flush out these toxins and algae. Natural occurrences such as droughts

also exacerbate the situation by allowing increased erosion. This introduces

particulates with adsorbed P into the water from surface runoff.

The farming fraternity, especially the pig, poultry and feedlot industries, tend to

spread their effluent in the neighbourhood of their facilities. With the animals in large

confined facilities, their sewage is often collected and partially treated. The treated

effluent is usually spread or pumped onto nearby land for disposal. Leaching of

soluble P is reduced if the disposal of effluent is performed on clay soils as the

3

‘phosphorus adsorption capacity’ of some clays can be extremely high. For

environmental reasons, phosphorus adsorption curves are used to determine a soil’s

capacity to adsorb phosphate and hence determine the suitability of soil for disposing

effluent onto the land (Ozanne and Shaw 1967; Rayment and Higginson 1992). The

focus of this research is the pig industry.

1.1.2 The Pig Industry

Monogastric animals such as pigs and poultry have a single stomach. Their diets are

regarded as a problem due to their inability to breakdown and absorb the P from

grains (especially the phytic acid), hence requiring P additives. The P that is not

broken down and/or digested by these animals will pass into the effluent. Effluent is

mainly composed of wastewater from manure, waste feed, shed cleaning, water spills

and leaks. Effluent collection systems are usually gravity fed pits and drains, with pre-

treatment of the effluent to remove large solids. The piggery effluent and solid is

usually applied to surrounding land by irrigation and spreading. Evironmental

concerns about pig effluent production and animal manure are the accumulation in

the soil of P, copper and zinc, and their high concentrations of nitrate and ammonia

(Jongbloed and Lenis 1998).

Piggeries produce over 100,000 tonnes of P per annum. Calculation of total P in the

effluent and solids is estimated by mass balance (APL 2004). If the available P

concentrations are incorrectly estimated then a much higher loading to the soil will

occur increasing leaching into the ground water and surrounding environment.

Nahm (2004) indicated a number of approaches to reduce potential P losses from pig

4

manure. They include reducing P solubility in manure, with chemical amendments

and fly-ash, and feed management practices such as the utilization of phytase,

addition of vitamin D metabolites and organic acids, maintenance of proper Ca:P

ratios with phytase and vitamin D or dietary manipulation. Nahm does not however,

indicate the available phosphorus concentrations and the importance they have in

the addition of inorganic phosphate sources, possibly because of the belief that they

are 100 % available to the animal. However, Nahm does say that the phosphate

requirements are met by addition of inorganic P minerals.

A number of different techniques have been used to reduce the amount of P in the

faeces. Those include the introduction of low phytic acid strains of grains, and the

addition of phytase to hydrolyze the phytic acid in the animal. Phytic acid is known to

be immobilised in clay soils as it is strongly adsorbed reacting with the metal oxides,

forming insoluble compounds (Turner and Leytem 2004). However, it was found to

be highly labile in sandy soils (Koopmans et al. 2003). Phytic acid is the major

organic component in pig and poultry faeces.

As a consequence, large amounts of phosphorus rich manure is usually applied to a

small land area, resulting in accumulation in the soil and the potential to cause

surface water contamination leading to algal blooms and eutrophication (NRC 1998;

Nahm 2004).

1.1.3 The project

This project was designed to investigate the management of phosphorus in feed

5

supplements used in intensive pig production facilities in Australia. It is intended to

assist in the development of economically and environmentally sustainable practices

in this key primary industry. Feed formulations include safety margins to reduce

deficiencies, but these will also cause additional waste problems. Improving

knowledge of feed ingredients will allow for the minimisation of these safety margins.

The availability of phosphorus in the inorganic phosphorus sources will be examined

in digestibility studies, where apparent digestibility (measuring the difference between

the uptake and output) of phosphorus will be used (CEFIC, 1999; Dellaert et al.

1990). Monocalcium phosphate has been included as a positive control as it is

considered that its phosphorus is completely available (NRC 1998) . “Palphos” is the

major rock phosphate used in Australia, while “Kynofos 21” is expected to be used

extensively as a commercial source of phosphate in Australia due to its balanced

Ca:P ratio.

1.2 Pig nutrition

1.2.1 The need for phosphorus

Phosphorus (P) is an essential nutrient (and macro mineral) in animals, comprising

over 25% of the body’s total mineral content. About 80% of the pig’s phosphorus is

found in bones and teeth. The remaining 20% is involved in numerous biochemical

reactions and biological products, ranging from energy production (e.g. ATP) to the

formation of cellular membranes (phospholipids and proteins) and nucleic acids (e.g.

DNA) (Cromwell and Cofffey, 1991; Kemme et al , 1994; NRC, 1998; CEFIC, 1999;

6

De Groote et al, 2003). Pigs absorb phosphorus from the small intestine in the form

of orthophosphate, with absorption predominantly occurring in the jejunum and

duodenum. Phosphorus deficiency may result in poor bone mineralization and

growth resulting in easily fractured bones. In the blood, inorganic P deficiency is

demonstrated by low concentrations and some elevated blood enzyme levels

(alkaline and calcium phosphatase) (Kopinski 2002a; Kopinski 2002b). For this

reason, over supplementation of P is traditionally used to maximize pig growth.

1.2.2 Amount of phosphorus needed

According to the Nutrient Requirements of Swine (NRC 1998), the total daily

phosphorus requirement for pigs with a body weight of 20 to 50 kg is 0.50 % of the

animal’s diet. This figure is derived from performance levels, age and genotype of the

animals. For P digestibility experiments, it is essential to select dietary P

concentrations that are marginally deficient, so as to ensure a linear response over

the range of P intake. The amount of the phosphorus absorbed is related to the

animal’s requirement.

The calcium to phosphate ratio (Ca:P) should be maintained within the recommended

range of 1:1 to 1.25:1 for total phosphorus, or 2:1 to 3:1 for available phosphorus

(amount utilised by the animal), to assist with growth. The calcium to phosphate ratio

used in this research was expected to be approximately 2.75:1 for the basal (the

basic feed mix used for all samples) and approximately 1.75:1 for the mixes with the

additives.

7

It is important to maintain homeostasis in the dietary supply of phosphorus in pigs.

Under normal conditions this is maintained by controlling the absorption rate of

inorganic phosphorus (Pi) in the upper small intestine and by renal Pi excretion.

These mechanisms are controlled by the parathyroid hormone (PTH) and calcitriol

(1,25-dihydroxycholecalciferol, 1,25-(OH)2D3). As the Pi concentration in the blood

plasma decreases, renal calcitriol production increases releasing more Pi from the

bones and soft tissues into the bloodstream. It also increases Pi absorption by

stimulating the intestinal tract to activate the sodium-coupled Pi-co-transport system.

As the Pi concentration increases in the plasma, hypophosphatemia will develop

suppressing the production of PTH and minimizing the loss of Pi from the urine. It is

important to maintain the proper Ca:P ratio, as hypercalcemia can develop from

increased intestinal Ca absortion and/ or Ca mobilization from the bone, also

suppressing PTH (Schroder et al. 1996).

1.2.3 Reduction of phosphorus excretion

Cereal grains and oilseed meal are usually the main ingredients in stock feeds and

phytate is the main form of P. Monogastric animals (pigs and poultry) are unable to

effectively utilise the P from these sources (NRC 1998; Kopinski 2002a; Bollinger et

al. 2004), because they lack the phytase enzyme to release the P. With only 25 to 40

% of the total P content of the grains used being available to the pig, P supplements

must be added to their diet, in order to meet the P requirements (Cromwell and

Coffey 1991; NRC 1998; Nahm 2004).

Extensive work has been done to increase P availability in monogastric animals

8

worldwide. The addition of the phytase enzyme into the feed mix, has been used to

assist in the breakdown of phytate. Vitamin D and its metabolites both stimulate the

phosphate transport mechanisms in the intestine increasing P absorption and

enhancing phytase activity, thus releasing more P from the phytate.

Phytic acid [myo-inositol 1,2,3,4,5,6-hexakis (dihydrogen phosphoric acid)] is an

important component of plants. The highest concentrations are found in seeds

where it usually occurs as mixed calcium – magnesium – potassium salts (and may

vary in structure slightly). Its main function is to store phosphorus and trace metals,

where the phosphate is released upon germination by hydrolysis, catalyzed by the

release of the enzyme phytase. Other forms of phytic acid are the scyllo-, D-chiro-,

and neo- inositol which are commonly found in soil (Turner and Richardson 2004).

Kemme et al (1999) suggested that the phytic acid was hydrolyzed by phytase to the

inositol forms hexa (6)-, penta(5)-, tetra(4)-, and tri(3)- phosphates. Kemme also

found that only the hexa form was found in the pig faeces. Hatzack et al (2001)

identified the different penta and tetra forms found in mutant forms of low phytate

maize and barley. The phytase enzyme hydrolyses the inositol hexaphosphate

liberating up to five inorganic phosphates and leaving inositol monophosphate.

Phytate can be regarded as an anti-nutritional factor as it can act as a chelator of

metal ions such as calcium, copper, magnesium and zinc, and can form crosslinks

with proteins and polysaccharides resulting in impaired protein and polysaccharide

digestion and nutrient deficiencies if the substances are bound. There are two

classes of phytases (3-phytase and 6-phytase) recognized by the International Union

of Pure and Applied Chemistry and the International Union of Biochemistry (IUPAC-

IUB). The differences are in the position of the phosphate on the phytic acid; the 3-

phytase removes the phosphate from position 3 first and the 6-phytase removes the

9

phosphate from the 6 position first. Successive dephosphorylations result in

intermediate forms including the free myo-inositol. 3-phytases are commonly found

in micro-organisms and filamentous fungi. The 6-phytase is commonly found in plant

seeds. Up to 88 % of total phosphorus in seeds can be in the form of phytin

phosphorus (Gibson and Ullah 1990).

Many companies will supply phytase in Australia; they include Alltech, BASF,

Bioproton and Roche. Even with the use of phytase, approximately 35 % of the

phytate P remains unavailable to the pigs fed on a corn-soy bean diet (Sands et al.

2001), thus reducing the amount of inorganic P required but not removing the

requirement for additional P completely.

Not only are researchers now adding phytase to assist in the digestion of phytate, but

a number of researchers are also exploring the use of genetically modified low-

phytate grain with similar amounts of total phosphate and showing large increases in

the bioavailability of P in these high available phosphorus grains (Spencer et al.

2000; Sands et al. 2001)

Although these techniques assist with the uptake of phosphorus there is still a

shortfall in the overall nutritional requirements for phosphorus thus requiring the

addition of feed additives usually in the form of inorganic sources. This is particularly

so for young animals and lactating animals as their requirements are higher for

phosphorus.

10

1.3 Feeds

1.3.1 Inorganic Phosphorus sources

Inorganic feed additives are essentially forms of inorganic fertilizers which are

suitable as additives in feed mixes, as they contain low concentrations of fluoride and

heavy metals. These low concentrations of fluoride and heavy metals will increase

the animal’s ability to utilize the nutrients in the mixture. A limited number of inorganic

P sources are available in Australia. There are mixtures of different phosphates, 4 to

5 sources of dicalcium phosphate from Europe and China and one presently,

acceptable source of rock phosphate (“Palphos”) available to the pork industry. A

wider variety of sources is available overseas and these are potential sources of

inorganic phosphorus in Australian pig diets.

The major source of inorganic phosphorus in Australia is tri-calcium phosphate or

rock phosphate. The other significant source of inorganic phosphorus has been

dicalcium phosphates. Generally, the phosphorus sourced from these products has

been considered to be completely available to the animal. With the concentration of

phosphorus added in excess of requirements in Australian feed, P availability was

never considered an important issue.

Phosphatic fertilizers are mainly sourced form phosphate rock, more commonly know

as hydroxyapatite (or hydroxylapatite), and commercially referred to as ‘tricalcium

phosphate’ (TCP): chemical formula [Ca10(PO4)6(OH)2, or 3Ca3(PO4)2.Ca(OH)2]. This

is the hydrated form of the apatite group of minerals, where the hydroxyl group (OH)

11

can also be replaced by fluoride (F) or chloride (Cl). The true form of tricalcium

phosphate [Ca3(PO4)2] is found naturally but may also be produced from the

hydroxyapatite by reacting it with phosphoric acid. Tricalcium phosphate is found in

two varieties, α- tricalcium phosphate [α- Ca3(PO4)2] and whitlockite [β- Ca3(PO4)2].

The other forms of calcium phosphates used are monetite (or dicalcium phosphate,

DCP), [CaHPO4], dicalcium phosphate hemihydrate, [CaHPO4.½H2O], brushite (or

dicalcium phosphate dihydrate, DCP), [CaHPO4.2H2O], monocalcium phosphate

(MCP) [Ca(H2PO4)2] and monocalcium phosphate monohydrate, [Ca(H2PO4)2.H2O]

(Halmann, 1972; Corbridge, 1978; De Groote et al., 2002). All feed phosphates used

in the European Union must conform to directives 2001/102/EC and 2002/32/EC on

undesirable substances. These directives specify that the maximum allowed limits

for the following substances are: fluorine < 2.0 mg/kg, cadmium < 10 kg/kg, arsenic

< 10 mg/kg, lead < 30 kg/kg, mercury <0.1 mg/kg and dioxins < 1 ng/kg as they can

cause animal and human health problems.

Dicalcium phosphate was commonly used in the feed industry for feed phosphate

that contains varying amounts of monocalcium phosphate, usually blended with

defluorinated phosphate sources or rock phosphate (Molins, 1991). This complicates

the problem of digestibility as the rock phosphate is regarded as less available by

some researchers (Mulder and Jongbloed,1985; Jongbloed, 1987; Dellaert et al

,1990; Kemme et al, 1994; Poulsen, 1995; De Groote et al, 2002; Frensenie et al,

2004 and Traylor et al, 2005).

12

1.3.2 Organic Phosphorus sources

The main forms of organic phosphorus are meat meals (MM), meat and bone meals

(MBM) and fish meals (FM). Teeth and bone contain high levels of calcium and

phosphorus, hence are commonly added to these organic feed additives. MBM is

produced by rendering, drying and grinding mammalian tissues and bone, with the

exclusion of hair, wool, hide, except where it naturally adheres to heads and hoofs.

These are usually recycled from leftovers from human consumption and are very high

in nutrients, including proteins, amino acids and minerals, in particular Ca and P

(MLA 2003). The main component of MBM is hydroxyapatite, it also contains other

forms of phosphorus, such as octacalcium phosphate and tetracalcium phosphate.

The forms of Ca and P found in the teeth and bone were reported by Halmann (1972)

to be much more digestible than the inorganic form of hydroxyapatite. The forms of

hydroxyapatite found in the teeth and bone are more commonly the carbonate form

with a small amount of fluoride. The carbonate substitutes for some of the

phosphates, and the extra positive charge is balanced by an extra hydroxyl or

fluoride (Halmann 1972). Bone apatite is also different from the inorganic form as it is

nanocrystalline, structurally disordered, and compositionally nonstoichiometric. It also

shows a high degree of variability among species, individuals and tissues (Cho et al.

2003). It must be noted that all the organic based feed additives must be analysed

on a per batch basis, as the nature of these samples mean the results will vary

between batches.

A ban on addition of animal meals into all diets for ruminants and other farmed

animals is in place in the European Union, and other countries due to cases of

bovine spongiform encephalopathy (BSE), also known as ‘mad-cow disease’, and the

13

link to the human variant of Creutzfeldt-Jacob disease (vCJD) (Pawitan et al. 2004).

BSE is a prion transmitted disease, and MBM is a possible source for transmitting the

disease. A transmissible spongiform encephalopathy (TSE) in sheep and goats,

Scrapie, the main suspect for BSE, is already common throughout the world (Heim

and Kihm 2003) and has been found in Australia. Bone meal and fat melting (mixed

animal fats) are also potential sources of the disease (Kamphues et al. 2001). This

has led to the investigation of more cost-effective ways of adding minerals such as

phosphorus in the formulations of fed animal diets (Sellier 2003). Australia is

currently regarded as BSE and TSE free and is rated by the European Commission

as category 1, the highest possible rating. Australia also has a ban on feeding

animal protein to ruminants, but still recommends the use of MBM for monogastric

animals, in aquaculture and petfood (MLA 2003). Traylor et al (2005) reported

increased performance from increased levels of MBM in swine diets, with high levels

requiring additional tryptophan.

The ban on addition of animal meal into all diets in the European Union is likely to

influence Australia in the next few years and our reliance on phytase and inorganic

sources of phosphorus is likely to increase. Although AUSPIG (2006) (APL’s

provider of information on feed ingredients, formulation diets, feeding levels and

other farming requirements) assumes that the availability of phosphorus from all P

sources is 100 %, there is evidence from Europe that the availability of phosphorus is

likely to range from 50 – 90 %. The possible increasing reliance on these inorganic

phosphorus sources to supply the majority of the phosphorus requirement of the

animal implies that we should accurately know the availability of phosphorus from

each of these sources. There are different techniques to determine the availability of

phosphorus from inorganic phosphorus sources. A chemical in vitro method (in

14

glass) that provides a straight forward method to accurately assess product quality in

the laboratory would be invaluable.

The question remains as to which P sources are more suitable for pigs from both an

economic point of view and an environmental point of view.

1.4 Analysis of digestability

1.4.1 In vivo Digestion

Total phosphorus (TP) in a diet clearly does not reflect its nutritive value (Jongbloed

1997; Jongbloed and Lenis 1998). Phosphorus ‘digestibility’ can be synonymously

used with ‘bioavailability’, ‘availability’, ‘apparent absorption’ and ‘retention’ all to

determine the amount of a particular nutrient utilised by the animal. Hence this topic

can be very complex, as there are several ways in which to define and determine the

digestibility. DPI Queensland (Kopinski 2002a) defines the following terms: TP is the

chemically determined total amount of phosphorus in a feed ingredient; available

phosphorus is the amount of phosphorus in a feed ingredient that can be utilized by

an animal. Availability is the proportion of the total amount in an ingredient that can

be utilized by the animal.

De Groote (2002) briefly describes the different terms (digestibility, absorbability,

(bio)availability and bio-efficacy) used to assess and express the nutritive value of

minerals for animals. Here digestibility and absorbability refer to the gastrointestinal

tract (feed – faeces). (Bio)availability is often misleading as it has several different

15

meanings, the National Research Council (NRC 1998) compares the nutritive value

against a reference assumed to be 100 % available, whereas the agricultural

research council (ARC 1981) describes it as the fraction retained in the body, feed –

(faeces + urine). Other distinctions are made between true and apparent, and ileal

and faecal.

Shen et al (2002) used a regression analysis technique on a corn based diet and

estimated endogenous P at 0.7 g/kg of dry matter intake. Ajakaiye et al (2003) who

used older and larger pigs than Shen et al, obtained 0.45 ± 0.21 g/kg of dry matter

intake with soybean meal. Pettey et al (2006) indicates the significance of

endogenous P, yet says that it is only a small percentage of the total P requirement.

Dilger and Adeola (2006) reported relatively low endogenous P, suggesting the

results were highly variable and not significantly different form zero. Edwards (1996)

regarded both endogenous P and urinary P excretion as not significant due to the

very low concentrations which are lost via these routes compared to that of faeces.

Pigs fed below the P requirements will excrete very little P in the urine (Jongbloed,

1987).

Tests for bioavailability of P in pigs can utilize a number of different factors. A list of

factors used are: the amount of bone ash, bone P concentration, P retained in the

empty bone and bending moment of several different bones (tibia/ toe/ metacarpal),

serum P, serum alkaline phosphatase, serum alkaline phosphate, growth

performance, urine phosphate content, phosphate retention (balance or slaughter

technique), phosphorus absorption (true or apparent), and phosphorus

absorption/digestibility (faecal or ileal). Estimates are all dependent on the response

criteria used.

16

Sands (2001) used plasma or serum P concentration due to the relative ease and

cost of collection, although it is less sensitive as a response criterion when

compared to bone criteria. The international association of the European (EU)

manufacturers of major, trace and specific feed mineral materials (Emfema) ranks

phosphorus absorption/digestibility (faecal) of the highest importance for assessing

the relative biological value of phosphorus sources in pigs using suboptimal supply

(De Groote et al. 2002).

1.4.2 In vitro Digestion

Fuller (1991) suggested the need for a better evaluation technique for the very

diverse range of raw materials available for inclusion in animal diets. He also stated

that the estimation of nutritive values should be separately assessed. In vivo methods

(i.e. within a living organism), requiring special housing facilities and equipment which

are expensive and slow. Also with the concern about using animals for routine

testing, in vitro methods which are rapid and reproducible are preferred. To be

successful, in vitro methods must be based on an understanding of the physical and

chemical composition of feeds and the digestive and metabolic processes in the

animal. Fuller (1991) lists five main areas to consider: the aim of the test, the in vivo

measurement with which it is to be compared, the need for a single test for all raw

materials for particular additives or classes of additives, the choice of appropriate

preparation/grinding for the test, and the requirement for accuracy as opposed to

estimates.

In Fuller’s book (Fuller 1991), Graham describes the design requirements of an in

17

vitro method. To predict the nutritive value of commercial feeds it must be accurate,

reproducible, rapid, cheap, simple, robust and adaptable, and if possible, results

should be additive, allowing values to be assigned to the additives / feedstuffs

irrespective of diet. Two main assumptions need to be made in the design: all soluble

material is digestible, and any anti-nutrient effects will often be lost in the in vitro

system. With all prediction methods, any in vitro method will only be as good as the

standards on which it is based.

The aim of the in vitro method is to mimic the processes which occur in the digestive

system, particularly that of the stomach which contains pepsin and HCl, and the small

intestine which contains bicarbonate and pancreatin. The pH at the start of the small

intestine, where the absorption of phosphorus occurs, is regulated to approximately

8.0. Yet the amount of calcium ions that form chelates with phosphates, increases as

the pH increases within the pH range of 8 to 11 (Molins 1991), rendering the P less

available at higher pH. Hence the in vitro methods using pH close to 8 are expected

to be the most reliable.

Spencer et al (2000) used an in vitro method described by Liu et al (1997) using HCl

and pepsin followed by bicarbonate and pancreatin in a dialysis tube, surrounded by

a succinate and NaCl buffer all at 39oC. They reported 57 % and 11 % respectively

for the low phytate corn and normal corn compared to the highly ranked metacarpal

bone bending moment data of 62 % and 9 % digestability.

The sample solution ratio used by others in vitro P techniques (Zyla et al. 1995; Liu et

al. 1997; Liu et al. 1998) is based on the digestive system and has a sample to

solution ratio of 1:2.65 or 1.0 g of sample with approximately 2 mL of hydrochloric

18

acid and 0.65 mL of sodium bicarbonate. Bollinger et al (2004) used sample weights

of 1.0, 0.5, or 0.25 g with a total volume of solution 5.3 mL. This sample to solution

ratio differs from the large dilutions used by the traditional methods used by the feed

companies, such as water soluble phosphorus (1 g of sample to 250 mL of water)

and 2 % citric acid soluble phosphorus (1 g of sample to 200 mL of 2 % citric acid) ,

which are based on methods used to analyse fertilisers. The methods based on the

digestive system also used dialysis to remove any soluble phosphorus, which will

also reduce the concentration of phosphorus in the solution allowing the equilibrium

of the phosphorus compounds/complexes to become soluble. These methods may

be used by researchers, they appear complex and difficult; attempts at simpler

techniques are reviewed in this research.

1.4.3 Spectroscopic techniques

Infrared (IR) spectroscopy is based on the principle that chemical bonds in molecules

absorb or emit infrared light when their vibrational state changes. IR spectroscopy

offers a potential alternative and a much faster and cheaper method for determining

apparent digestibility. Most large livestock feed companies already use IR

spectroscopy to predict factors in cereals and grains such as fibre content, protein,

moisture, digestible energy content, and some even use it for amino acids (Van

Barneveld et al. 1999; Van Kempen 2001), and in meat and fat for colour, fat quality

and content and palatability (texture and flavour) (Van Kempen 2001) . Another

advantage is that it requires minimal or no sample preparation. No reagents are

required and hence no waste is produced, and it has the potential for on-line use and

the simultaneous evaluation of different properties. Disadvantages are its

19

dependence on reference methods, laborious calibrations, weak sensitivity to minor

constituents and limited transfer of calibrations between different instruments

(Prevolnik et al. 2005).

Near infrared reflectance spectroscopy (NIR) has been extensively used for the

analysis of animal feeds, grains and many other products. The main strength of NIR

compared to the other IR regions is that it has very low noise. However, a typical

spectrum of a solid animal feed has a baseline gradually increasing (as the

wavelength increases) as a result of scattering of light from the particles. NIR peaks

are indistinct and usually difficult to assign to specific components (Qiao and Van

Kempen 2004). NIR can provide compositional and other related information on the

sample with little or no sample preparation and no generation of chemical waste

(Reeves III 2001). Calibration of the spectral information requires multivariate data

analysis such as partial least squares (PLS) regression to relate to the product of

interest (Esbensen; Thermo-Galactic 2003). First or second order derivatives of the

spectra are often used for the PLS regression because they tend to improve the

calibration (Qiao and Van Kempen 2004).

Few minerals have been determined by NIR as there are few spectral absorptions in

this range and they are usually correlated to organic components in the diet. The

nature of NIR absorbtions are limited to those from CH, NH, and OH groups and are

due to overtones (large changes in vibrational state) and combination bands.

Reeves III (2001) and Bertrand et al (2002) used NIR and Mid infrared reflectance

spectroscopy (MIR) to determine minerals in samples, these include Ca, P, Fe, Al, K,

Mg, Mn, S, Cu and Zn as totals or extractables.

20

MIR consists of fundamental absorptions (primary vibrations) due to the CH, NH, and

OH groups and many others including inorganic groups , yielding sharper and more

clearly defined peaks(Reeves III 2001; Van Kempen 2001). Both Reeves III (2001)

and Van Kempen (2001) state that results from MIR calibrations are often more

accurate than those based on the NIR and so they are better suited to quantitative

purposes. NIR are recorded on a wavelength domain, whereas MIR spectra are

recorded in the frequency or wavenumber domain. Reeves III (2001) attempted the

analysis of minerals including total Ca and P, by both NIR and MIR. The NIR had a

high R2 for Ca calibrations but with a large standard deviation, especially at the lower

concentrations. Calibrations for the other analytes were very poor and therefore not

a satisfactory replacement for conventional assays. MIR had almost identical results

for Ca, and poor results for all other analytes with some improvement from the NIR

which, while not satisfactory for total Ca, showed some potential for determining P.

Mid-infrared diffuse reflectance (MIR-DRIFT) spectroscopy coupled with partial least

square (PLS1) has been used in this study to predict the apparent digestibility of

phosphorus in particular feed additives. Compared to previous techniques, MIR does

not require dilutions with substances such as KBr. Bertrand et al (2002) indicated

that the dilution of these solid samples is likely to cause problems which will interfere

with the spectra. Noise is regarded as a problem for MIR because the sample

penetration by the light source is a function of wave number, with deeper penetration

at higher wave numbers. Hence removal of spectra at lower wave numbers will

remove spectral data which is more prone to noise from shallower sample

penetration (Qiao and Van Kempen 2004).

Partial Least Squares (PLS1) is a form of spectral decomposition which uses the

21

constituent information during the decomposition process, allowing the spectra

containing higher constituent concentrations to be weighted more heavily than those

with low concentrations. It performs the decomposition on both the spectra and the

constituents simultaneously. This takes advantage of the existing relationship

between the constituent and the spectra. PLS1 is a predictive technique which

computes the optimal linear relationship between a high number of correlated

variables (from the spectra) and a limited number of observations (the results). The

PLS1 technique is designed to get as much constituent information as possible into

the first few loading vectors, which include a loading for each reflectance value

(eigenspectra).

Spectral data can be analysed by score. The score is the result of one loading vector

plotted against another loading vector. When all the samples used to calibrate are

scored (known as a training set) it should form a cluster in the plot. Cluster analysis

is used to check for samples with inconsistent scores based on their separation from

the cluster. This can be done by measuring the Mahalanobis distance. The distance

is calculated from the sample point to the cluster mean. The distance is also scaled

for the range of variations in the cluster in all directions and assigned a probability in

terms of standard deviation. An outlier is a sample with a Mahalanobis distance

larger than 3 standard deviations (Thermo-Galactic 2003).

The number of factors or loading vectors is determined for each model. As each

factor is calculated, it is ordered by constituent weighted variance. The factors will

eventually start to contribute very small amounts which may be regarded as

system/instrument noise. Usually 3 – 7 factors will account for 99 % of the variation.

The method used to determine the best number of factors is called Prediction

22

Residual Error Sum of Squares (PRESS). With this Cross-Validation of the PRESS

emulating the prediction of “unknown” samples by using the training set data, starting

with the factor counter at 1. The point at which the graph reaches a minimum and/or

starts to ascend again is regarded as the optimum number of factors to use (Thermo-

Galactic 2003), the graphs of PRESS versus Factor Number can be seen in

Appendix H.

Calibration by PLS1 begins with the ‘leave one out’ cross-validation procedure. As

each new factor is calculated, the spectral concentration scores are switched before

the contribution of the factor is removed from the raw data. The resulting data are

then used to calculate the next factor. This process is repeated until the

required/requested numbers of factors are calculated. The PLS1 algorithms for

quantitation of the eigenvectors and scores are fairly complex. The following

references provide further information (Haaland and Thomas 1988; Bertrand et al

2002; Thermo-Galactic 2003).

NMR spectroscopy is widely used in chemistry, although the most common

experiments are performed on liquid samples or extracts. As 31P is the only naturally

occurring P isotope (100% natural abundance) all P species within a sample can

potentially be detected by this technique. Solid- State 31P NMR spectroscopy is a

powerful tool and has been used to characterize phosphorus in many medical,

environmental and agricultural samples (Condron et al. 1997). However, it is limited

by low natural concentrations of P. As the concentration of P in feed samples is high,

this should not pose a problem. Also spectral resolution is poor compared to solution

spectra. Anisotropic interactions are both an advantage and a disadvantage, but will

reduce the signal and broadening the peaks in a solid. The anisotropic interactions

23

are usually not available for samples in solution, due to the continuous rapid motion

of nuclei in a liquid. Many other possible techniques such as measurement of inter-

nuclear distances, torsion angles, atomic orientation, spin diffusion, molecular

dynamics and exchange processes are also available in solid state NMR.

One of the problems with using extractions on agricultural samples being prepared

for analysis, is the potential to change the compound being analysed. When pH is

varied, the degree of protonation of P compounds will change and the observed NMR

chemical shift will be the average of the contributions from differently protonated

forms of the same P compound (Koopmans et al. 2003; Cade-Menun 2005). Hence

the solvent may be an important determination of observed chemical shift

differences. This may effect the choice of extractions, as for agricultural samples

(soils and manure) there is usually a 4 or 5 part extraction, giving an equal number of

fractions. First, a water and a bicarbonate extraction is used to assess the labile

species. Then sodium hydroxide is used to extract iron and aluminum bound species

and occluded organic phosphate species (higher concentrations of hydroxide can

cause hydrolysis (Leinweber et al. 1997; Koopmans et al. 2003)). This step may also

include a chelating agent such as EDTA for the extraction of monoesters such as

phytic acid (Koopmans et al. 2003; Turner 2004; McDowell and Stewart 2005; Toor et

al. 2005) and finally hydrochloric acid is used to dissolve the calcium phosphate

phases. Hunger et al (2005) pointed out that the P compounds in each of these

extractions were not specific to that extract. Toor et al (2005) recognized the inability

of the extractions to identify the inorganic P components, it only allows for

quantification of total inorganic P. Due to the complexation of P with paramagnetic

cations such as Fe, Al and Mn, the inorganic P compounds will all appear as one

orthophosphate NMR peak. Koopmans (2003) indicated the possibility that some

24

transformation of organic P to inorganic P cannot be excluded, due to alkaline

hydrolysis of orthophosphate diesters with the use of sodium hydroxide extractions.

Calcium phosphate was present in all extractions, while aluminum was present in the

hydrochloric acid extract hence caution is required if employing these extraction

techniques due to the effects caused by the paramagnetic cations. McDowell and

Stewart (2005) indicated the possibility of movement of the available P into

recalcitrant pools during drying of the manure, hence making it less soluble and

changing the speciation.

For solid samples at the magic angle of θ = 54.74o, the resonance frequency of the

crystal is not altered by the heteronuclear dipolar coupling ((3cos2θ – 1) = 0). By

taking advantage of this angle, and orienting the internuclear vector at this angle, the

dipolar coupling is zero. This technique is known as Magic Angle Spinning (MAS)

(Jerschow et al. 2002). The spinning speed is also important as on decreasing the

speed more spinning sideband peaks will appear. An external standard of 85 %

H3PO4 is used.

The dominant P compounds expected in faeces are the inorganic complexes of Ca

and Mg with some Al and organic compounds such as phytic acid. Iron and

manganese phosphates are not expected to be detected due to paramagnetic line

broadening (Hunger et al. 2004). Cade-Menun (2005) reported chemical shifts for

some P compounds from solid state 31P NMR studies: these shifts are at 9 ppm for

dicalcium phosphate dihydrate, 3 ppm for hydroxyapatite, -2 ppm for monetite (or

MCP) and -5 ppm for crandallite [(CaAl3(OH)5(PO4)2]. Other studies show dicalcium

phosphate at -1.5 or 0 ppm and 1.7 ppm for the dihydrate (Hunger et al. 2004).

Unfortunately, these and other references all have peak positions that appear to

25

differ even though the samples types should be similar and hence give the same or

similar peaks.

Toor et al (2005) indicated that changes in animal diets can alter the chemical

composition of manure and the speciation of manure P, and in turn may affect the

potential for P retention or loss in manured soil. It was also suggested that total P

analysis in manure for the addition to soil is not appropriate as it does not indicate the

availability to the soil or the possibility of run off. Hunger et al (2004) states that the

dicalcium phosphate (DCP) and calcium carbonate contained in the feed are

dissolved during digestion and will be excreted in some form, while phytic acid is

known to be found in poultry manure where grain has been added to their diet.

Solid state NMR is not suitable for the determination of organic compounds, due to

both peak broadening (causing overlap) and the low concentration levels usually

present. Hence for organic compounds, the use of liquid NMR is preferred, while for

inorganic compounds the solid state NMR is preferred.

1.5 Research outline

1.5.1 Rationale for research

The major source of inorganic phosphorus in Australia is tri-calcium phosphate or

rock phosphate (CEFIC 1999; De Groote et al. 2002). The other significant sources

of phosphate have been from meat meal and dicalcium phosphate. Generally, the

phosphate sourced from these products has been considered to be completely

26

available to monogastric animals and with the level of over formulation of phosphorus

in Australia, availability was never considered an important issue. The increasing

attention to phosphorus pollution from intensive animal agriculture has paved the way

for products such as phytase to be incorporated into Australian pig diets to improve

the uptake of phosphorus from phytate sources (ASoE 2001; Nahm 2004). An

alternative strategy to the inclusion of phytase to reduce the amount of phosphorus

excreted into the environment is to reduce the total phosphorus in the diet by using

more highly available sources of inorganic phosphorus. The ban on addition of

animal meal into all diets in the European Union is likely to influence use in the next

few years and our reliance on phytase and inorganic sources of phosphorus is likely

to increase (Heim and Kihm 2003; Sellier 2003). Current evidence from Europe and

the rest of the world shows that the availability of phosphorus in these organic

sources will range between 50 – 80 %, not 100 % as was originally believed and

reported by AUSPIG (2006). The increasing reliance on these inorganic phosphorus

sources to supply the majority of the phosphorus requirement of the animal implies

that we must accurately know the availability of phosphorus from each of these

sources. There are different techniques to determine the availability of phosphorus

from inorganic phosphorus sources. A quick and easy method such as an in vitro

chemical method or a method that provides a way to accurately assess product

quality in the laboratory would be invaluable.

Extensive research has been conducted in the Netherlands to examine the P

availability of several types of feed phosphates from different origins (Mulder and

Jongbloed 1985; Dellaert et al. 1990). In addition, different techniques have been

used to determine the biological availability of P and Dellaert et al. (1990) concluded

that the apparent digestibility of P is the most efficient criterion to evaluate the

27

nutritional value of various feed phosphates. Stein et al (2008) showed that the

digestibility of calcium and phosphorus in mono calcium phosphate is not influenced

by the concentration of either calcium or phosphorus in the diet. Stein et al (2008)

also reported that as phosphorus in the diet increased so did the phosphorus

absorbed and retained by the animal giving a linear relationship. Any new data will

be useful to the pig feed industry.

1.5.2 Objectives:

The research was designed to provide valuable information on the formulation of

diets for intensive pig farming operations and in particular to:

• determine the apparent digestibility of phosphorus from major inorganic

sources of phosphorus available to the Australian Pork industry.

• develop and compare various in vitro chemical methods to assess the

availability of phosphorus in inorganic phosphorus sources.

• recommend an efficient, accurate in vitro chemical method that has potential

to assess the availability of phosphorus in inorganic phosphorus sources.

This project will provide industry with data on the availability of phosphorus in several

common sources of inorganic phosphorus. This information can be used in diet

formulation for the pork industry to ensure that sufficient phosphorus is supplied to

28

meet the nutritional needs of pigs while ensuring that excess phosphorus is not

excreted to contaminate soil, and potentially waterways. Chemical methods of

assessing phosphorus availability have been developed and compared for their

ability to provide a quick assessment of phosphorus digestibility.

The hypotheses examined in this research are:

Apparent digestibility of phosphorus in dicalcium phosphate, monocalcium

phosphate, combination phosphate and rock phosphate is 100 %.

The apparent digestibility of phosphorus in inorganic phosphate sources can be

accurately predicted from an in vitro chemical test.

1.5.3 The project description

The apparent digestibility of phosphorus in six common sources of phosphorus (five

inorganic P sources and one meat and bone meal) was determined in a metabolism

study using 8 pigs with a live-weight of 40 – 80 kg. Their starting weight was

between 40 to 45 kg and final weights between 70 and 80 kg. A basal diet was

formulated to contain no inorganic phosphorus source and was found to have 0.31 %

total phosphorus. To this diet 0.17 % phosphorus was added from one of the six P

sources to create the other six experimental diets. This gives a total phosphorus

concentration of 0.48 % which is still below the requirements stated in the Nutrient

Requirements of Swine (NRC 1998) of 0.50 %. This will ensure the highest possible

absorption of phosphorus by the pigs (Jongbloed 1987; Potter et al. 1995).

29

The animal research involved the use of an 8 by 4 randomised digestibility

experiment. This required the use of two basal controls and six inorganic feed

formulations allocated over four different runs. The apparent digestibility of

phosphorus between each phosphorus source was calculated by difference (input

minus output). A series of in vitro chemical tests were used to determine the

solubility of P.

It has been specified that pigs require a preliminary feeding period of 7 to 10 days to

completely remove any residue not from the experimental diet (Schneider and Flatt

1975). If this period is observed the marker to marker approach is not required.

Schneider and Flatt explain the many different types of markers used as well as their

advantages and disadvantages. A preliminary period of 9 days was chosen as it is

essential to feed the animals on the same weighed feed for several days before

collecting any faeces. This is done so that the content of the digestive tract may

become free of any material originating from any previously ingested food. Each run

comprised a total 14 day feeding cycle, consisting of 9 days preliminary feeding and

a final 5 day period for total faeces collection where the 8 pre-fed pigs had been

weighed and held in metabolism crates. The total length of the pig trial was 56 days

1.5.4 In vitro methods chosen

Three tests were based on extractable and exchangeable nutrient analysis

techniques, used to interpret the availability of soil nutrients (in particular phosphate -

phosphorus) (Rayment and Higginson 1992). A fourth test involved an initial

extraction with hydrochloric acid followed by a pH adjustment to 8 with bicarbonate

30

and hydroxide to mimic the digestive tract, but using a higher extraction ratio and

without dialysis to speed up the reaction.

0.01 M calcium chloride is regarded by some soil analysts as a “universal” extractant

and is commonly used in phosphate sorption analysis or buffering capacity (Ozanne

and Shaw 1967; Rayment and Higginson 1992). Calcium chloride induces

precipitation of dissolved organic matter and associated P, especially if high

molecular weight compounds are present (Koopmans et al. 2003). In this instance

only the MBM contains organic matter which may be a problem.

Ammonium acetate is also regarded as one of the “universal” extractants, but is more

commonly used for the extraction of cations in soil (Rayment and Higginson 1992) to

determine their availability to plants. Due to its use for the exchange of cations, such

as calcium, it should be able to liberate the phosphate ions from the calcium ions,

assisting in analysis. Acetate is similar to citrate in its nature and so should have a

similar effect, although the pH of the extractant solution will also influence the

solubility of the phosphate. Acetate was also used by Morgan (1941) for the analysis

of phosphate in soil.

The use of bicarbonate solution is very popular in soil science for the determination of

available phosphorus, as demonstrated by the many methods which use this

extractant with different extraction times and ratios (Olsen et al 1954; Colwell 1963;

Rayment and Higginson 1992). It is also the major component in bile from the

intestinal tract and has been used by other researchers (Zyla et al. 1995; Liu et al.

1997; Liu et al. 1998; Spencer et al. 2000; Bollinger et al. 2004) for the analysis of

available phosphorus.

31

Non destructive techniques, such as infrared (IR) analyses (both NIR and MIR) were

used to look for differences between the sources. Solid state magic angle spin

nuclear magnetic resonance (SS-MSA-NMR) was used to examine differences

between the samples and to determine if there were any changes in the form of P

passing through the animals. A statistical package was used to determine the

relationships between various in vitro chemical and IR information and in vivo

digestibility. Due to the variability between sources and between batches, an

additional ten samples were collected from suppliers and analysed by the chemical

and IR methods which showed the greatest promise as predictors of P digestibility.

32

Chapter 2

Materials

And

Methods

33

2.1 Collection of the feed additives

The initial stages of this research involved the collection and sub-sampling of various

feed additives. The additives selected were commonly used as sources of

phosphorus. These additives were essentially forms of inorganic fertilizers which

were suitable as additives in animal feed mixes.

Figure 1. Brands of feed additives used in the trials

The brands of feed additives used in this project (Figure 1) were: rock phosphate or

tricalcium phosphate (Palphos, TCP), dicalcium phosphate (Rediphos, DCP),

Chinese dicalcium phosphate (Honghe P/L, DCP), European dicalcium phosphate

(Bolifor, DCP dihydrate), mono/dicalcium phosphate mixture (Kynofos 21, MDCP)

34

from South Africa, and meat and bone meal (QAF limited, MBM, Australia formally

know as Bunge). These materials were sourced from different feed and fertiliser

companies, with the majority supplied by QAF limited.

2.2 Testing of the Feed

Samples from each of the additives given to the pigs were analysed using DPI

Werribee’s Inorganic Chemistry Method 00092 “Determination of calcium,

magnesium, phosphorus, potassium, sodium and sulphur in Fertilisers, Soil and

Sediment by Block Digestion and Inductively Coupled Plasma Emission

Spectroscopy”. Method 00092 is based on a nitric – perchloric acid digestion (using

block digestion in test tubes with small funnels to reflux the acid) followed by analysis

using Inductively Coupled Plasma Emission Spectroscopy (ICP-ES). For each

sample a mean concentration and standard deviation was calculated from triplicate

analyses. The method was based on ‘Official Methods of Analysis (1995) 16th edition

and the AOAC International Methods 929.02, 942.02 and 957.02 (AOAC, 1995).

2.3 Preparation of the feed

In preparing the feeds, the total phosphorus value was kept at below the 0.50 %, set

as the mimimum requirement by the Nutrient Requirements of Swine (NRC 1998) to

ensure the highest possible absorption of phosphorus by the pigs which weight

between 20 and 50 kg. The amount of total phosphorus aimed for in the basal diet

was approximately 0.29 %, each of the six prepared feeds required an additional

0.17 % total phosphorus from its associated additive to produce an aimed target of

35

0.46% total P. The calcium concentration is also used with the amount of

phosphorus to keep the calcium to phosphorus ratio within the recommended range

of 1:1 to 1.25:1 for total phosphorus, and 2:1 to 3:1 for available phosphorus, for

optimum phosphorus absorption. The calcium to phosphate ratio used in this

research was expected to be approximately 2.75:1 for the basal and 1.75:1 for the

mixes containing additives.

Following the anaylsis of the different feed additives and ingredients, the program

called ‘FEEDMANIA DIET’ was used to calculate the amount of different additives

required and to balance the diet so that all other nutritional requirements were met, at

the cheapest possible price. A sample of the basal diet calculation from the printout

obtained after all the relevant data had been added into the program can be found in

Appendix B. The other six diets were very similar to the basal diet with the exception

that they contained the particular additive to make total phosphorus in the feed mix

approximately 0.17 % w/w higher. The Feedmania results (Appendix C) show that

the differences between the diets were slight. Main differences were in the amount of

starch, the volume of sunflower oil (measured by weight) which is used mainly as a

lubricant/wetting agent for the feed, the amount of limestone to balance out the

amount of calcium in the total mix, and the individual phosphorus additives being

tested. Due to there being the equivalent of two pigs fed on the basal diet for every

pig fed on one of the other six diets, double the amount of feed was needed for the

basal diet, hence the requirement of 300 kg of feed instead of 150 kg as per the other

diets. Table 1 indicates the actual amounts of feed ingredients calculated for use to

make up the total feed mixes, after rounding for accuracy in the balances available.

The Basal Diet or base diet is generally the principle diet without any of the additives

being tested. It has minimal or no influence on the animal’s ability to digest the

36

compound being tested. No makers were added to the diets.

Table 1. Weight used for individual feed ingredients (kg) as determined by ‘Feedmania’.

Diet Number 1 2 3 4 5 6 7

Ingredient

Basal

Palphos

Rediphos

DC

P

Honge

DC

P

Bolifor D

CP

Kynofos

QA

F MB

M

Corn 214.5 107.5 107.5 107.5 107.5 107.5 107.5

Soybean meal 37.50 18.75 18.75 18.75 18.75 18.75 18.75

Blood meal 9.00 4.50 4.50 4.50 4.50 4.50 4.50

DL methionine 0.174 0.087 0.087 0.087 0.087 0.087 0.087

L. lysine 0.546 0.273 0.273 0.273 0.273 0.273 0.273

Mineral Vitamins 0.600 0.300 0.300 0.300 0.300 0.300 0.300

Salt 0.600 0.300 0.300 0.300 0.300 0.300 0.300

Starch 21 11.5 10.5 10.5 10.5 10.0 4.64

Sunflower Oil 6.00 3.00 3.00 3.40 3.00 3.00 3.00

Limestone 9.97 2.349 3.411 3.183 3.429 4.055 2.802

Palphos 1.485

Rediphos DCP 1.395

Honge DCP 1.362

Bolifor DCP 1.391

Kynofos 1.208

QAF MBM 8.05

Total Weight (kg) 300 150 150 150 150 150 150

37

Table 2 shows the balances used at the feed shed in the pig research training centre

(PRTC) compound with their weight ranges and their accuracy. Once the different

mixes had been determined, they were then prepared using balances provided

(Table 2) to the achieve ingredient weights as close as possible to those in Table 1.

All the ingredients for each particular diet were placed into a large mixing container

and mixed for a minimum of ½ hour. A sub sample was then taken for storage in a

freezer (stored by DPI Werribee 600, in case samples required re-analysis) and

another one was taken for analysis (to DPI Werribee 621) using DPI Werribee’s

Inorganic Chemistry Method Number 00092. The results are given in Appendix C.

Table 2. Balances and their accuracy

PRTC Balance No. Weight range

(g)

Accuracy

(g)

20 1000 ± 0.5

12 100 ± 0.05

14 5 ± 0.001

The individual diet mixes were then transferred into the labelled feed hoppers (refer

to Figure 2). Once the feed had been mixed they were then ready to feed to the pigs.

38

Figure 2. Feed hoppers containing the feed mixes for this project.

2.4 The animal testing

Individual pigs were labelled as in Table 3 and in Figure 9, i.e. 22/7, 7/9, down to

21/8. These labels were given to the pigs by the PRTC. The numbers 1 to 7 refer to

each diet type as shown in Table 1. The pigs were then relabeled alphabetically for

easier identification.

The preparation of the animals for testing required the formation of an 8 by 4

randomised digestibility experiment (Table 3). It was conducted at the Victorian

Institute of Animal Science (DPI Werribee 600), with two control basal diets (diet 1)

and 6 inorganic phosphorus sources in each run. The latter were rock phosphate

(Palphos, diet 2), Dicalcium phosphate (Rediphos, diet 3), Chinese dicalcium

phosphate (Honghe P/L, diet 4), European dicalcium phosphate (Bolifor, diet 5),

mono/dicalcium phosphate mixture (Kynofos 21, diet 6) and meat and bone meal

(QAF limited, diet 7). These were compared to the two control basal diet experiments

39

in each run, over 4 separate runs. The allocation of the diets to the pigs themselves

was randomised. Table 3 shows the randomised selected diets and which pigs were

given the different diets (from Table 1).

Table 3. Randomized Allocation of Diets

Pig ID PRTC ID Run 1 Run 2 Run 3 Run 4

A 22/7 2 6 1 4

B 7/9 1 5 3 2

C 10/7 6 1 7 3

D 11/5 4 2 1 7

E 22/10 5 7 5 1

F 11/4 7 3 6 1

G 8/7 1 4 2 5

H 21/8 3 1 4 6

2.5 The Selection of the Pigs

To achieve the 8 by 4 randomisation experiment, ten grower pigs were selected from

the pig research training centre (PRTC at Werribee) herd at about 40 kg live weight

(average weight being 42 ± 1 kg (mean ± SD)) and moved into individual pens for

observation. Of these ten animals, eight were selected at random. The two extra pigs

were to be used as replacements in case of illness, injuries, or factors affecting

animal appetite, none of which occurred. On 4 occasions at 14 day intervals the 8

pre-fed pigs were weighed and moved into the metabolic crates for a 5 day period for

total faeces collection. They were also weighed when moving out of the metabolic

40

crates. The total length of the pig trial was 56 days.

2.6 Feeding of the animals

Feed level was set at 1.5 kg per day and this stayed the same as the pigs grew in all

four of the runs. Water was provided from nipple waterers. Wet feed refusals (only

one incident) were placed in a pre-weighed tray and re-weighed. Metabolism crates

were set-up with a faeces tray and a feed spillage screen. The spillage screens were

covered with pre-weighed and numbered Chux wipe cloths on which spilt feed

accumulated. At the end of the 5 day collection period, the Chux were oven dried

(100 oC) then allowed to equilibrate at atmospheric conditions before re-weighing, to

determine the amount of spilt feed.

2.7 Collection of faeces and urine

Every second Monday to Friday the total faeces collected were added into a pre-

weighed plastic bag (1 for each of the 8 pigs) weighed, and stored in a freezer. After

the collection on Thursday, the frozen samples in the bags were allowed to thaw,

then after the last collection for the period on Friday the faeces were mixed (see

Figure 3 for a photo of the mixing of the faeces), as per the PRTC procedure. Three

sub samples were collected from the homogenized faeces, one was frozen as a

backup, and a second was used to determine moisture or dry matter content of the

faeces in triplicate. The last sample was also oven dried at 100 oC. Dried faeces

were then ground in preparation for analysis and placed into a 250 mL jar. Urine

samples were collected from each of the 8 pigs on the Wednesday of each collection

41

period. Urine samples were then frozen. A drinking water sample was also collected

and frozen. Urine was not collected quantitatively as it was used only to confirm

deficiency of phosphorus, although boredom can led to excessive drinking and

urination. The drinking water sample was tested to determine the possibility of

phosphorus entering the animal through the consumption of the water. The faeces,

urine and water were then analysed using DPI Werribee’s Inorganic Chemistry

Method Number 00092. The analyses of the urine samples were taken to confirm that

the dietary P concentrations were below requirement. It is expected that pigs fed

below the P requirements will produce urine containing less than 150 mg P/L

(Jongbloed, 1987).

Figure 3. Mixing bowl used for homogenising the faeces.

42

2.8 Testing Sites

The majority of the testing and experimental work was done at the DPI Werribee

Centre. Handling and mixing of the feed was at 600 Sneydes Road Werribee, and

the in vivo (animal) testing and housing of the animals was also carried out at the Pig

Research Training Centre (PRTC) 600 Sneydes Road Werribee. The majority of the

in vitro testing, with the exception of the NIR and Solid State-Magic Angle Spinning-

NMR (SS-MAS-NMR) was performed at 621 Sneydes Road Werribee. The NIR was

performed at DPI Horsham, and the SS-MAS-NMR was conducted at CSIRO

Clayton.

2.9 Calculation of P- Digestibility

The digestibility coefficient is described as the percentage of a nutrient consumed /

ingested (% input), which does not appear in the faeces (% output). The digestibility

of the feed is calculated by difference. For this research the nutrient is P, but this

calculation can be used for any nutrient.

Each pig was fed a weighed amount of feed, which was used to calculate the mass of

each nutrient eaten, rejected or spilt. The feed mixes were then analysed, as fed or

as received, determining the amount of nutrient within the mix, then corrected to %

for dry matter (A = mass of nutrient per mass of dry food). The amount of feed eaten

by the animal is calculated by the amount of feed offered (FO) minus the amount of

feed refused or left uneaten (R), minus any feed spilt (S). The feed spilt was collected

using a ‘CHUX wipe’, which was dried prior to using the mass in the calculation as it

can also contain spilt water and saliva. All calculations were done on a dry weight

43

basis [Dry Matter (DM)], the S value in Equation 1 has been corrected for the weight

of the CHUX wipe which was originally in a dry weight basis. Equation 1 shows the

formulae used to calculate the total nutrient intake (I) for the animal. All weights are in

grams.

Equation 1.

Total Nutrient Intake (I)

I = [(FO – R) x DM /100] – S x A

I = Total nutrient intake (g)

FO = amount of feed offered (g)

R = amount of feed refused (g)

DM = % dry matter of food

S = amount of spilt feed (g)

A = % mass of nutrient in the food corrected for dry matter

The next step was to calculate the output (O), or the amount of nutrient found in the

faeces. The faeces were collected (F), (in this case collecting faeces over five days

decreased the errors associated with irregular excretion from the animal) and dried

(Dry Matter (DM), again at 100oC). The container set faeces were weighed and the

weight of the container (a large plastic bag) was subtracted. All weights were

measured in grams. The dried faeces were homogenized, and analysed for the

nutrient (N) content, which will be P, using DPI Werribee’s Inorganic Chemistry

Method Number 00092. The digestion of the faeces required the addition of hydrogen

peroxide to the nitric acid digestion prior to the addition of perchloric acid (as

perchloric acid can explode in the presence of organic matter). Water and urine

44

samples were analysed in duplicate by this method for the majority of the samples.

However, the contribution from these sources was not significant and therefore not

included in the equations. The Output was calculated from Equation 2.

Equation 2.

Total Nutrient Output (O)

O = (F x DM /100) x N

O = total nutrient output (g)

F = faeces collected (g)

N = % mass of nutrient per mass of dry faeces

After obtaining the total nutrient intake and total nutrient output from the animal, the

percentage of digestibility in the test diet (Digtest as shown in Equation 3) was

determined. The difference between nutrient intake and the output, or nutrient which

did not come out in the faeces (and /or urine), is regarded as nutrient which has been

digested.

45

Equation 3.

Digestibility in the Test Diet (Digtest)

Digtest = 100 x (I – O) / I

Digtest = % digestibility in the test diet

I = Total nutrient intake (g)

O = total nutrient output (g)

2.10 Digestibility by difference

Once the digestibility of each test diet had been determined, the digestibility of the

nutrient additives could be determined. This determination is known as digestibility by

difference, or the difference between the Basal diet and a mixture of the Basal diet

with additional nutrient additives. Therefore two trials were required to determine the

digestibility of a nutrient additive, one to determine digestibility in the basal diet and

the other to determine the combination of the basal diet with the additive. An

assumption is made that the digestibility of the basal diet will not be altered by the

additive; thus any additional nutrient in the faeces must come from the additive. It is

also assumed that the endogenous P excretion (any loss of P from other forms; hair,

skin, sweat, urine, etc) is very low and similar for all diets, so is not significant.

Therefore the digestibility of the additives can be determined by difference.

From the digestibilities of the test diets, the digestibility of the basal diet (DB) was

determined, and so was the digestibility of a basal plus additive diet (DBA). The

percent of nutrient within the mix (A) previously determined, was re-labeled to

46

differentiate between the basal (as AB) and the basal plus additive (as ABA) diets. By

subtracting these two values (ABA – AB) the nutrient concentration in the additive

(AI) was obtained. This should correlate with the estimated nutrient concentration of

the individual ingredients of the feed. The digestibility by difference of the nutrient in

the additive must therefore be the converted back to the nutrient concentration.

Consequently the Digestibility must be the total nutrient concentration in the basal

diet plus additive (ABA) times the digestibility of that nutrient in the basal diet plus

additive (Digtest ABA) minus the nutrient concentration of the basal diet (AB) times the

digestibility of that nutrient in the basal diet (Digtest AB), the result will be divided by

the nutrient concentration in the additive (AI or ABA – AB), resulting in Equation 4.

Equation 4.

Digestibility of the nutrient in the additive.

Digadd = (ABA x Digtest ABA – AB x Digtest AB) / AI

Digadd = % digestibility of the nutrient in the additive

ABA = % nutrient concentration in the basal diet plus additive

Digtest ABA = % digestibility of that nutrient in the basal diet plus additive

AB = % nutrient concentration of the basal diet

Digtest AB = % digestibility of that nutrient in the basal diet

AI = % nutrient concentration of the additive

47

2.11 Invitro testing

In vitro (from the Latin ‘in glass’) is regarded as a man-made form of testing. These

experiments attempt to duplicate the digestibility of P in the additive by techniques

which do not include animal testing or any part of the animal. The technique is much

cheaper and easier to reproduce.

2.12 Methods for testing

The methods that were selected for testing include both wet chemical techniques,

and non-destructive forms of testing. The wet chemical methods include 7 different

extraction types which incorporated variations in the sample weight to extractant

volume ratios, in an attempt to improve the efficiency of the extractions. The

extraction matrixes used were: water, calcium chloride, citric acid, ammonium

acetate, sodium bicarbonate and hydrochloric acid. The non-destructive techniques

used were near infrared (NIR), mid infrared (MIR) and Solid state NMR spectroscopy.

2.13 Wet Chemical Methods

The following set of extracts (including the water soluble phosphorus and 2% citric

acid extracts), were all measured by the SKALAR SAN PLUS segmented flow auto-

analyser system due to its wide dynamic range, with the ability to pre-dilute and post-

dilute extracts. The extracts or supernatant were measured colorimetrically using the

molybdate blue complex (Fogg and Wilkinson 1958; EPA 1979) measured at 880 nm.

A dialysis step was used for sample clean-up preventing large molecules (including

48

coloured compounds) from entering the determinative stream and potentially causing

interference. Dialysis also effectively minimises troublesome CO2 evolution when a

bicarbonate solution is used in acidic conditions. Phosphate reacts in the presence of

ammonium molybdate and potassium antimony tartrate to form a phospho-antimony-

molybdate complex. This complex is reduced by ascorbic acid and produces an

intense blue colour. The intensity of the colour is measured at 880 nm with

background correction at 1010 nm and directly correlates to the amount of

phosphorus in the solution. Silicates, arsenic and any other interferences can

generally be ignored. The calibration standards used were equivalent to those in the

DPI Werribee’s Inorganic Chemistry Method Number 00120 “Determination of the

phosphorus adsorption capacity of soils by automated flow colorimetry” with the

exception of the matrix matching solution, in which the standards and the rinse

solutions were made. For alkaline matrices such as the sodium bicarbonate, the

method recommends the use of 1 M sodium hydroxide solution instead of a 4% v/v

solution of sulphuric acid.

2.13.1 Chemicals and reagents

Reagents

• Deionised water (18.2 M Ω cm-1, “Milli-Q”)

• AR Grade BDH Potassium dihydrogen orthophosphate KH2PO4

(Dried 2 hours at 105 oC prior to use)

• AR Grade BDH Sulphuric acid 18 M, H2 SO4

• AR Grade BDH Ammonium molybdate tetrahydrate,

(NH4)6Mo7O24.4H2O

• AR Grade BDH Antimony potassium tartrate, K(Sb)C4H4O6.5H2O

49

• AR Grade BDH Ascorbic acid, C6H8O6

• Skalar FFD-6 wetting agent

• AR Grade BDH Calcium chloride, CaCl2.2H2O

• AR Grade BDH Sodium hydrogen carbonate, NaHCO3

• AR Grade BDH Sodium hydroxide, NaOH

• AR Grade BDH Ammonium acetate, CH3COO.NH4

• AR Grade BDH Ammonium hydroxide, NH4OH

• AR Grade BDH Glacial acetic acid, 99 % w/v CH3COOH

• AR Grade BDH Hydrochloric acid 10M, 32 % w/v HCl

• AR Grade BDH Citric acid, C6H8O7.H2O

• Merck 1000 mg/L P standard.

Working reagents

Solutions were prepared using A grade volumetric glassware.

Standards

• Stock solution 1000 mg/L P

A 4.3935 g mass of dried potassium dihydrogen orthophosphate was weighed into a

1000 mL volumetric flask, dissolved in approximately 700mL of deionised water and

made up to one litre.

• Intermediate stock solution 100 mg/L P

A 10 ml aliquot of stock solution (1000 mg/L P) was diluted into a 100 mL volumetric

flask and made up to volume with deionised water.

50

• Calibration standards

Calibration standards were prepared as outlined in DPI Werribee’s Inorganic

Chemistry Method Number 00120 “Determination of the phosphorus adsorption

capacity of soils by automated flow colorimetry”

The concentrations were 0.5, 1.0, 2.0, 5.0, 10.0 and 20.0 mg/L of P as PO4.

• Secondary standard

A purchased phosphate standard (Merck 1000 mg/L of P) was diluted to 2.0 mg/L of

P as PO4 to bring the final concentration within the range of the calibration standards.

Solutions

• Sodium hydroxide solution

A 4 g mass of sodium hydroxide was weighed into a 1000 mL volumetric flask,

dissolved in approximately 700mL of deionised water and made up to one litre, then

2 mL of FFD6 was added and the solution mixed.

• Sulphuric acid solution

To approximately 700 mL of deionised water, a 40 mL aliquot of sulphuric acid was

added then allowed to cool and made up to 1 L before 1 mL of FFD6 was added and

the solution mixed.

• Distilled water + FFD6

A 1 ml aliquot of FFD6 was added to 1 L of deionised water.

51

• Ammonium molybdate solution

To approximately 700 mL of deionised water, a 40 mL aliquot of sulphuric acid was

added, and a 4.8 g of ammonium molybdate was weighed into the flask. The

solution was allowed to cool and made up to 1 L, then 1 mL of FFD6 was added and

the solution mixed.

• Stock potassium antimony tartrate

A 1.5 g mass of potassium antimony tartrate was weighed into a 500 mL volumetric

flask, dissolved in approximately 300mL of deionised water and the solution made up

to 500 mL.

• Ascorbic acid solution

An 18 g mass of ascorbic acid was weighed into a 1000 mL volumetric flask. This

was dissolved in approximately 700mL of deionised water, and 20 mL of the stock

potassium antimony tartrate was added and the solution made up to one litre with

deionised water.

Extractants

• 0.01 M calcium chloride

A 1.470 g mass of calcium chloride was weighed into a 1000 mL volumetric flask.

This was dissolved in approximately 700mL of deionised water and made up to one

litre.

52

• 0.5 M sodium bicarbonate (pH 8.0)

A 42.0 g mass of sodium bicarbonate was weighed into a 1000 mL volumetric flask.

This was dissolved in approximately 800mL of deionised water and approximately 0.4

g of sodium hydroxide was added to adjust the pH of the solution to 8.0 before

making up to one litre.

• 1.0 M ammonium acetate

A 77.0 g mass of ammonium acetate was weighed into a 1000 mL volumetric flask.

This was dissolved in approximately 700mL of deionised water, adjusted to pH 7.0

with ammonium hydroxide or glacial acetic acid and made up to one litre.

• 0.01 M hydrochloric acid

A 1.0 mL aliquot of 10 M hydrochloric acid was added to approximately 700 mL of

deionised water and made up to 1 L

• 0.7 M sodium bicarbonate extracting solution

A 58.8 g mass of sodium bicarbonate was weighed into a 1000 mL volumetric flask.

This was dissolved in approximately 800mL of deionised water and made up to one

litre.

• Water

Deionised water (18 M Ω cm-1, “Milli-Q”)

• 2 % Citric Acid

A 20 g mass of citric acid was weighed into a 1000 mL volumetric flask. This was

dissolved in approximately 800mL of deionised water and made up to one litre.

53

2.13.2 Procedures

The following tests were selected on the basis of their potential to indicate the

digestibility of P in the feedstock. The first 5 tests involve sample extraction. The

rationale for each choice is described briefly below.

The extraction volumes and feedstock mass for test 1 to 5 are presented in Table 4.

Test 6 involved washing the samples through a filter paper. It was chosen because it

is commonly used in the feed industry as a guide for P digestibility.

Test 1: 0.01 M calcium chloride

This extraction was selected due to its common use as an extractant for soils. It has

the ability of saturating the cationic phase with calcium and settling out fine

particulate suspensions after equilibrium.

Test 2: 0.5 M sodium bicarbonate with different extraction times.

The bicarbonate extraction method is commonly used for estimating the available

phosphorus in soil (Rayment and Higginson, 1992), and with methods such as Olsen

et al (1954) and Colwell (1963) phosphorus. The bicarbonate solution was buffered

to pH 8.0 (instead of pH 8.5 as used in the soil methods) with sodium hydroxide,

which is more representative of the pH found in the ileum, where phosphorus is

absorbed by the pig. Two different extraction times and volumes were used.

54

Test 3: 1.0 M ammonium acetate

The ammonium ion (NH4+) exchanges places with cations such as calcium, and the

acetate ion (CH3COO-) (like the citrate ion used with fertilisers and feed additives)

assists in the extraction process by forming an ion pair with the cations. This allows

P to become ionised in solution and available for colourimetric analysis.

Test 4: hydrochloric acid and sodium bicarbonate

This technique is designed to closely resemble the two sequential extractions that

occur in the digestive system (without enzymes) prior to the adsorption of

phosphorus by the pig. Four different sample solution ratios were used.

This was done as a two step extraction starting with initial extraction with 0.1 M

hydrochloric acid solution for the appropriate time. The second extractant of 0.7 M

sodium bicarbonate was added slowly directly to the initial extractant (pH was then

adjusted with sodium hydroxide so the blanks gave a pH of 8.0). Frothing was

allowed to subside prior to analysis.

Test 5: 2 % citric acid

The 2 % citric acid extraction is based on ‘Directive 77/535/EEC, method 3.1.3’ which

has been superseded by ‘Regulation (EC) No 2003/2003 method 3.1.3’ and is used

by the IFP (Inorganic Feed Phosphates) Quality feed phosphates. It can be found

under ‘inorganic feed phosphate test methods’ at

http://www.feedphosphates.org/guide/test_method.html#solubilitycitricacid

55

Test 6: water soluble phosphorus

This test is based on the AOAC Official Method 977.01 Phosphorus (Water-Soluble)

(AOAC, 1995; Halmann, 1972) in Fertilizers (a 1:250 sample to solution ratio). A 1.0

g mass of sample was weighed into a folded 110 mm No.5B Advantec filter paper

and the paper and sample were placed onto a funnel on a 250 mL volumetric flask.

The sample was then washed thoroughly with small portions of deionised water.

After each portion the sample was allowed to drain prior to adding the next portion.

This was continued until the collected filtrate in the volumetric flask was over 200 mL.

The flask was then made to volume and a portion was taken for analysis by the

molybdate blue method, with standards made up in deionised water.

56

Table 4. Test Method weights, volumes and extraction times.

Test Extractant

used

Sample

Mass (g)

Extractant

Volume (ml)

Shaking

Time (hours)

1 A 1.000 10 1

2 B 1.000 20 0.5

2 B 1.000 10 16

3 C 1.000 20 0.5

4 D 0.250 10 + 30* 1 + 1*

4 D 0.500 10 + 30* 1 + 1*

4 D 0.750 10 + 30* 1 + 1*

4 D 1.000 10 + 30* 1 + 1*

5 E 1.000 200 2

* Extractant D consists of a two stage extraction which requires the addition of two separate

extractant volumes and extraction times.

Extractants:

A 0.01 M calcium chloride

B 0.5 M sodium bicarbonate at pH 8.0

C 1.0 M ammonium acetate at pH 7.0

D 0.1 M hydrochloric acid then 0.7 M sodium bicarbonate

E 2 % citric acid

The sample mass which (as shown in Table 4 was recorded to 3 decimal places) was

weighed into a 50 mL plastic centrifuge tube, or a 250 ml plastic bottle for test 5.

Extractant solution was then added to each container which was capped and shaken

to thoroughly wet the sample. The bicarbonate solution (extractant B and the second

57

part of extractant D) was added slowly and in small volumes to allow the frothing

reaction with the acidic matrix to subside. The containers were placed in an end-over-

end shaker for the selected shaking time(s) at 18 rotations per minute. After

extraction with the citric acid (test 5) the solution was immediately filtered through a

folded 110 mm No.5A Advantec filter paper, discarding the first 20 mL of filtrate. All

the other solutions were centrifuged at 3000 rpm for 10 minutes. A sub sample was

removed, appropriately diluted and analysed for P on the auto analyser.

2.14 Non-Destructive Methods

These techniques offer important advantages over wet chemical methods. They

require minimal or no sample preparation, and no reagents (and hence no waste is

produced). They can also be very cheap (especially compared to the in-vivo

techniques) and are non-hazardous. The infrared techniques are quick and easy to

use, especially once the calculations have been determined. Infrared detectors are

often found in feed mills, thus allowing the feed ingredients and completed feeds to

be monitored onsite(Van Kempen 2001).

Infrared (IR) Spectroscopy

IR spectroscopy measures the absorbance due to molecular vibrational and

rotational energy. The fundamental ranges for the spectral regions in nm

(wavelength) are: visible 380 – 780, near IR 780 – 3000, mid IR 3000-30,000 and far

IR 30,000 – 300,000 or in cm-1 (wavenumber) visible 26320 - 12820, near IR 12820 -

3333, mid IR 3333 - 333, and far IR 333 – 33.3. In near IR spectroscopy large

changes in vibrational states are observed (overtones), while in the mid IR region

58

changes primary vibrations state are detected. Thus mid IR yields sharper, more

clearly defined peaks that are better suited to quantitative analysis.

Infra red spectra were analysed using the chemometric software packages, METLAB,

WINISI II (Infrasoft international.LLC) (for the NIR), and GRAMS-AI 7.02 (Thermo

Galactic) with partial least squares (PLS1) attachment (for the MIR). The analysis of

the spectra used models employing PLS1 regression with internal cross-validation.

The optimal number of PLS1 factors were chosen according to the difference in

successive cross-validation errors. As the predictive ability of IR spectroscopy

depends largely on the reliability of reference material, the results from the in-vivo

digestibility tests and the total phosphorus analyses were used as the internal

references for this purpose.

NIR (near infrared)

Two spectrophotometers were used : the Foss/Tecator NIR systems 6500 and NIR

systems 5000 utilising the reflectance radiation from the solid samples: the spectral

wavelengths run between 400 and 2500 nm which includes the visible segment (400

– 750 nm), the Herschel segment (750 – 1100 nm) and the NIR segment (1100 –

2500 nm). Reflectance data of the entire spectrum, as log 1/R (reflectance) were

collected at 2 nm intervals

Samples were placed into the sample holder (rotating sample cups), the holder was

capped, and then placed into the instrument sampler, where the sample was rotated

during collection of the spectrum.

59

MIR (mid infrared)

The instrument used was a Perkin-Elmer (P.E.) Spectrum One FT-IR Spectrometer

using diffuse reflectance. The infra red spectra ranging between 7800 to 450 cm-1 (or

1280 to 22,220 nm) at 8 cm-1 resolution were collected for all the feed additives.

Finely ground samples were gently pressed into the sample holder, leveled and then

placed into the MIR where h the acquisition of each spectra took 1 minute.

Analysis of the spectra (both the NIR and the MIR) was done using the GRAMS-AI

PLS software (PLSPlus IQ). This software is designed to build calibration models

using multi-variate methods, such as partial least squares regression analysis. Using

the full spectrum, these techniques were fitted to multiple data points to improve

sensitivity. The software may select some of the hundreds of co-linear or orthogonal

variables within the spectrum.

For the MIR each sample was analysed a minimum of seven times. From this set of

data two spectra from each sample were selected to produce a prediction curve

using the GRAMS software. The spectra were mean centered and, using the partial

least squares (PLS1) software, calibration equations were developed with both raw

and first derivative data. The optimum number of PLS1 factors were determined

using cross validation with one file removed.

Solid State 31P Magic Angle Spinning NMR

Solid state NMR, allows for the direct examination of samples with minimal

preparation or alteration and requiring only a small sample size (Finely ground

samples were used in this experiment). A Varian 300 Unity Plus, 300 MHz wide bore

magnet operating at a magnetic field of 5.28 Tesla corresponding to resonance

60

frequencies of 121.4 MHz was used for 31P. Spinning speeds were maintained at 7

to 10 kHz. An external standard of 85 % H3PO4 was used and the chemical shifts

were referenced against it and set to 0 ppm. Standard parameters for a 90o pulse

were, PW = 4.5 µs, SW = 40 kHz, NS = 32, DI = 60 sec, DI ≥ 5.T1 and Decoupling

ON, where

• PW: the pulse width (the time the pulse is turned on).

• SW : the spectrum width determines the highest frequency that needs

to be sampled.

• NS: the number of acquisitions/spectra taken.

• DI: the time between pulses allowing for full decay.

• T1: the time required for acquisition.

Figure 4 shows the NMR pulse sequence, illustrating the pulse width, the acquisition

time, and the pulse delay for two subsequent pulses.

Figure 4. NMR pulse sequence

Solid state NMR was used for the analysis of the different diets, sample ingredients

and the faeces. It was used to predict what ingredient had passed through the pig

without change and to determine if there was a difference between the inorganic

additives, especially the different DCPs.

61

Chapter 3

RESULTS

&

DISCUSSION

62

3.1 Outline

This chapter presents the results and a discussion of in vivo and in vitro studies

conducted to determine the phosphorus digestibility of pig diets prepared from a

range of commercial ingredients. The studies were designed to allow accurate

analysis of digestibility of the diets using live pigs and to subsequently compare the

data with data obtained from simulated in vitro digestibility studies which could

provide a simpler and less expensive means of optimising pig diets.

The total calcium (Ca) and phosphorus (P) concentrations were estimated in each of

the diet ingredients used in the experimental diets and these were used to determine

the diet mix using the program, FEEDMANIA. The amounts in each diet were aimed

to achieve approximately 0.8 % Ca and 0.46 % P (0.29 % for the basal diet). Each

diet was then tested to confirm the actual amounts of Ca and P (Section 3.2.2).

The diets in this trial should not be used for any other purpose as they were

specifically designed to be deficient in P for determining the digestibility of the

ingredients being tested, not as a replacement feed. Each of these ingredients being

tested are commonly found in the feed industry as Ca and P supplements, especially

for monogastric animals, due to their inability to properly digest phytate as it is found

naturally in grains.

Analysis of urine can indicate whether the pigs are deficient in P allowing maximum

absorption of P. Weight increase would show if the pigs are healthy and growing with

the diets given, and absorbing P from the diets. Their faeces were analysed in order

to estimate the P digestibility. Comparisons were made between the in vivo

63

techniques and the wet chemical techniques, and then against the spectroscopic

techniques, including a comparison with the estimated results for the total Ca and P

analysis.

3.2 The diet and its ingredients

3.2.1 Ingredients

The feed ingredients were analysed for Ca and P to accurately determine the

quantities of the different ingredients needed to be added into each diet mixture. The

results for Ca and P are shown in Table 5. Since they were not a focus of this study,

the results for other analytes, such as magnesium, sodium, potassium and sulphur in

each of the feed mixtures are provided in Appendix A. Once these quantities were

determined by the FEEDMANIA program, the ingredients were weighed out as

previously described on page 38 in table 1 (results shown in Appendix C), into a large

mixer and then mixed together.

64

Table 5. Calcium and phosphorus concentrations in the feed ingredients.

Calcium Phosphorus Ingredients of Feed

(%w/w) SD* (%w/w) SD*

Ca:P

Ratio

Rediphos - DCP 24.5 0.5 18.3 0.6 1.34:1

Honghe - DCP 28.8 0.4 18.7 0.4 1.54:1

QAF – MBM 5.9 0.3 3.2 0.2 1.8:1

Palphos – TCP 39 1 17.2 0.3 2.2:1

Kynophos – MDCP 16.8 0.3 21.1 0.3 0.80:1

Inor

gani

c fe

ed A

dditi

ve fo

r

indi

vidu

al d

iets

Bolifor – DCP 24.3 0.4 18.3 0.3 1.33:1

Blood meal 0.058 0.003 0.206 0.007 0.28:1

Maize 0.006 0.001 0.28 0.01 0.02:1

Limestone 21.8 0.9 0.012 0.002 1800:1

Soybean meal 0.42 0.03 0.69 0.07 0.61:1

Ingr

edie

nt u

sed

in a

ll di

ets

Vitamin mineral 9.2 0.3 0.042 0.003 220:1

*SD standard deviation in %w/w of three replicate samples.

Each of the inorganic feed additives are simple phosphate salts such as: whitlockite,

brushite, and monetite which are natural forms of calcium phosphate. Based on the

Ca and P concentrations of the ingredients supplied (as specified by the product

labelling) the type of salt in each of the products can be identified. Palphos contains

38.6% Ca and 17.2 % P, which is consistent with TCP which has 38.6 % Ca and 20

% P. Rediphos and Bolifor both contain approximately 24.5 % Ca and 18.3 % P,

which consistent with DCP dihydrate which has 23.3 % Ca and 18.0 % P and

Honghe contains 28.8 % Ca and 18.7 % P, which is consistent with DCP which has

29.5 % Ca and 22.8 % P. Kynophos 21 contains 16.8 % Ca and 21.1 % P, which is

predominantly in the form of mono calcium phosphate (MCP) as a monohydrate

65

which has 15.9 % Ca and 24.6 % P and mixed with a DCP; this is also confirmed by

its acidic nature. The nature of the salts in each case were confirmed by their

percentage of Ca and P (Table 5), after taking into account the moisture content. The

Ca and P concentrations in the sample of meat and bone meal (Diet 7) were

significantly lower (approximately 40 percent lower) than accepted published values;

meat and bone meal (MBM) is often considered to contain about 10 % Ca and 5.0 %

P (SCA, 1987; Kopinski, 2002a) . Analysis of a second supplied source of MBM also

had low Ca and P concentrations indicating care must be taken when using different

sources of additives for feed, as the actual nutritional results may not reflect those

published, especially if taken from a biological process. The Kynophos 21 labeled as

21 % P was shown to be correct, but not all the bags were labeled with nutritional or

elemental information, and the Palphos had been subsampled into a different bag

prior to receipt for the experiment. The concentrations of P in DCP is variable, as

shown by the difference between the two natural DCP products (Brushite and

Monetite), suggesting that nutritional labelling (concentrations of Ca and P) is very

important for feed additives, not just labelling the product as DCP.

All calculations were performed on a dry weight basis [Dry Matter (DM)] for all

ingredients, as shown in Table 6. These samples were received in a dry powdery

state. The much higher moisture content of the Bolifor DCP indicated that this

ingredient consisted of a hydrated form (or forms). The most common hydrate for

DCP is the dihydrate form, most possibly brushite, which has similar Ca and P

concentrations. Stoichiometrically there should be approximately 17 % water for the

dihydrate form of DCP, or 83 % DM, which is similar to the 85.5 % DM determined for

the Bolifor.

66

Table 6. Dry matter content, dried at 100 deg C for the P-sources. P-Source DM

% w/w

SD*

Palphos TCP 99.96 0.02

Rediphos DCP 98.22 0.03

Honge DCP 94.3 0.3

Bolifor DCP 85.51 0.07

Kynofos MDCP 94.02 0.05

QAF MBM 93.8 0.1

*SD standard deviation in % w/w moisture of three replicate samples.

The FEEDMANIA program was used to calculate the aimed amounts of 0.80 % w/w

for Ca in all diets, 0.29 % w/w for P in the basal diet, and 0.46% w/w P in the other

six diets. The amount of P in the basal mix was manually calculated from the

concentrations and weights used for the ingredients (including the vitamins, blood

meal, corn, and soybean meal) to check on the FEEDMANIA program. The result

obtained was 0.289 % w/w which confirmed results of the FEEDMANIA program.

The MBM diet was calculated to add 0.172 % w/w P to the total diet and the other P

sources were all estimated to add 0.170 % w/w to their diets. The Ca in all diets was

calculated to be within 0.803 % w/w and 0.805 % w/w. None of these calculated

results were significantly different from the original estimates from FEEDMANIA. The

experimental error is similar to the error expected from the variability in the balance

used.

67

3.2.2 Dietary mixtures.

The feed mixtures were then analysed to determine the total amount of nutrient in

each mixture (Table 7 and Appendix C). The analysed values were slightly higher for

total P than the calculated values. Table 7 also shows the variability in the Ca

concentration of 0.09 %w/w from the calculated concentration of 0.80 % w/w. The P

concentrations were all within 0.02 % w/w of the calculated concentration except for

the QAF-MBM diet which was 0.07 % w/w higher. The mean and standard deviation

of the inorganic feed concentration were 0.47 ± 0.01 % w/w and 0.75 ± 0.03 % w/w

for P and Ca respectively. The calcium to phosphate ratio used in this study was

2.4:1 for the basal feed and a range of 1.5:1 to 1.7:1 for the additives (see Table 7).

Table 7. Calcium and phosphorus results for each diet.

Phosphorus Calcium Diet

No. Diet type

%(w/w) SD* %(w/w) SD*

Ca:P

Ratio

Diet 1 Basal 0.31 0.01 0.74 0.03 2.4:1

Diet 2 Palphos 0.48 0.01 0.75 0.02 1.6:1

Diet 3 Rediphos DCP 0.46 0.01 0.74 0.04 1.6:1

Diet 4 Honge DCP 0.47 0.01 0.71 0.03 1.5:1

Diet 5 Bolifor DCP 0.48 0.02 0.80 0.01 1.7:1

Diet 6 Kynophos 0.48 0.02 0.76 0.03 1.6:1

Diet 7 QAF_MBM 0.53 0.01 0.89 0.03 1.7:1

*SD standard deviation in % w/w of three replicate samples.

*Results were obtained in triplicate using DPI Werribee’s Inorganic Chemistry Method 00092 which is

based on a nitric – perchloric acid block digestion

68

The results of Ca and P analysis for each diet (Table 7) collected during the pig

digestibility study revealed close agreement to the expected values based upon

analysis of the ingredients prior to mixing. The major exception was diet 7, (MBM)

where actual concentrations of P were approximately 10% higher and Ca was

approximately 20% higher than the respective expected values. The extra ten to

twenty percent does not make up for the significantly lower than expected Ca and P

results for the MBM. The higher results are most likely due to the heterogeneity of

the QAF-MBM (Diet 7), and the inability to take a relatively reproducible sub-sample.

Triplicate analyses also showed large variations as indicated by the standard

deviation of 0.3 % P and a mean of 5.9 % w/w P giving a CV of 5 % (Appendix A).

The analysis of the MBM also had the highest concentrations of potassium, sodium

and sulphur. Diet 7 did not show any significant difference in its variance when

compared to the other diets.

The Basal diet also gave a higher than expected result of 0.31 % w/w P (Appendix D)

compared to the calculated value of 0.29 % w/w P. This is approximately 7 % higher

than expected; while this may not be significant it highlights the amount of error that

can be produced from these mixes (which are not always homogeneous), and the

importance of accurate measurements.

The concentration of P in the basal diet was found to be 0.31 % w/w. The addition of

a further 0.17 % w/w total P from one of the six feed additives should have yielded a

total P concentration of 0.48 % w/w. This concentration is still below the Nutrient

Requirements of Swine (NRC 1998) of 0.50 % w/w. This deficiency ensures

absorption of P will be linear (Potter et al. 1995; Jongbloed, 1987). The MBM diet

actually yielded a value of 0.53 % w/w which is slightly above the requirement of 0.50

69

% w/w total P. The alternative to total P requirement is the available phosphorus

requirement of 0.23 % w/w (NRC, 1998). This requirement was also satisfied with a

concentration of 0.22 % w/w (Hastad, 2004). The average digestible P of the MBM

diet result was still above the minimum requirement. The results for the total P in the

urine samples for diet 7(MBM) indicated P was still deficient. This was due to the

amount of feed available to the pigs being below the recommended requirement,

thus bringing the overall level below requirements.

Table 8. Dry matter content of each diet, dried at 105 deg C.

Diet DM

% w/w SD*

1 87.5 0.1

2 87.5 0.1

3 87.3 0.2

4 87.3 0.1

5 87.2 0.1

6 87.2 0.2

7 87.9 0.2

*SD standard deviation in % w/w of three replicate samples.

The dry matter (DM) content of each diet was determined in order to calculate the

digestibility of the P in the ingredient tested. The NRC (1998) based its dietary

requirements on a DM of 90 % w/w for its diets. It also presumes corn and other

grains to have similar DM values. The results shown in Tables 7 and 8 provide the

information required to calculate the total nutrient intake using Equation 1 (page 46).

The results for the total nutrient intake are listed later in Table 11.

70

3.3 The Analysis of Urine and Faeces.

The analysis of the urine was used to determine if the pigs were deficient in P and if

the concentrations were high (compared to the faeces), they could be used in the

calculation for the digestibility of P. The analysis of the faeces was used in the overall

calculation of the digestibility of P.

3.3.1 The Analysis of P in Urine and Water

-30.0

-20.0

-10.0

0.0

10.0

20.0

30.0

40.0

50.0

60.0

1 2 3 4 5 6 7

Diet Type

Con

cent

ratio

n (m

g P/

L)

Figure 5. Total phosphorus concentration (mg/L) in pig urine for each diet

type (1 to 7)

The concentrations of P in the urine samples (Figure 5) of pigs fed the experimental

diets were very low (6 ± 7) mg/L (mean ± SD) with the maximum value of 40 mg/L

compared to 150 mg/L which is the maximum concentration observed in pigs fed

below the P requirement in other research (Jongbloed, 1987). These results

suggest that the diets would elicit the highest possible absorption of P by the pigs

from their feed, and would show a linear correlation.

71

Diet 6 (Kynofos) was skewed by an outlier of 40 mg/L where the other three runs all

yielded results for P of approximately 1 mg/L. In this run, the sample also had high

values for salts and other analytes. This may indicate possible dehydration of the pig

at the time of sampling. It may also have been caused by poor kidney function.

Edwards (1996) reported that urinary P excretion contributed less than 2 % of the

total P excreted in normal diets. This research had an average of 5.5 mg/L of total P

in the urine compared with approximately 31,000 mg/kg of total P in the faeces. The

total mass of faeces for the test animals ranged between 1 and 2.5 kg wet weight

over a full 5 day period, with a daily average of 0.31 kg (or 0.13 kg dried). Urine was

estimated at between 1 to 3 kg per day averaging around 2.5 kg per day. Therefore

P loss through the urine per day was approximately (5.5 mg/L x 2.5 kg) 0.014 g of P

(assuming a solution density of 1.0 g/cm3). On the other hand, dried faeces

contained an average of 3.1 % total P equivalent to (31 g/kg x 0.13 kg) 4.0 g of total

P per day. This equates to 0.34 % of the total P coming from the urine. This value is

significantly less than 2 % of the total P excreted (as determined by Edwards (1996)),

which is still much less than the variation within the faeces. Thus the total P of the

urine, in this thesis, can be ignored in the calculation for available P.

A drinking water sample was also analysed, with a result of < 0.5 mg P /L (similar to

the sample blanks) indicating there was no additional P in the diet due to water

consumption.

72

3.3.2 The Analysis of Faeces

Total faeces were collected over a five day period, mixed, dried (Dry Matter (DM))

and finely ground, and then re-homogenised prior to analysis for nutrients (as per the

instructions in the PRTC (pig research and training centre) and ICS (inorganic

chemistry section) methods). Figures 6 and 7 show the % nutrient (Ca and P

respectively) found in the faeces. The results for these are also shown in Table 9.

The total nutrient output is calculated from Equation 2 (results can be reviewed in

Table 12) shown in section 2.9. The digestibilities of the additives were determined

by difference. It was assumed that the digestibility of the basal diet would not be

altered by the additive, or differences in the diet types, hence any additional nutrient

in the faeces must come from the additive.

3.0

3.5

4.0

4.5

5.0

5.5

1 2 3 4 5 6 7Diet Type

Ca c

once

ntra

tion

(% w

/w)

Figure 6. Total calcium concentration (% w/w) in pig faeces for each diet

type (1-7)

The basal diet (diet 1) result was expected to produce a concentration of much lower

P in the faeces, as it was originally 0.17 % w/w lower in the diet. The Basal diet

73

result was the lowest recorded but statistical analysis showed that the result was not

significantly different from any of the others (except the Palphos, in diet 2). This

reveals the inability of the pig to digest P from certain components in the basal diet

mix, one component of which is known to be problematic, is the phytate. As the

phytate should be equal in all diets, it was expected that a similar amount of phytate

P would pass through each pig into the faeces. Phytate is regarded as the major

form of organic P found in faeces, and Kemme et al (1999) stated it remained

unchanged in the hexa- form. Thus any additional amount of P would be from the

inorganic additives.

2.4

2.9

3.4

3.9

1 2 3 4 5 6 7Diet Type

P co

ncen

tratio

n (%

w/w

)

Figure 7. Total phosphorus concentration (% w/w) in pig faeces for each

diet type (1-7)

As all the feed mixes started with the same amount of Ca they were expected to all

have similar results in the faeces. Figure 6 demonstrates a much higher Ca output

for the Palphos (diet 2) when compared to the other diets. The P (Figure 7) output

was also higher than the other diets for Palphos. This indicated that the solubility of

P in Palphos (diet 2) was lower than in the other diets. It also indicated the inability of

the pig to completely digest the Palphos in order to absorb the nutrients (Ca and P)

74

from it.

Table 9. Calcium : Phosphorus Ratio in pig faeces for each diet type.

Diet type

Basal

Palphos

Rediphos

DC

P

Honge

DC

P

Bolifor

DC

P

Kynofos

QA

F MB

M

Diet Number 1 2 3 4 5 6 7

Calcium in Faeces

( % w/w) 3.9 4.7 3.6 3.6 3.9 3.9 3.7

Ca SD* 0.4 0.7 0.2 0.4 0.4 0.4 0.4

Phosphorus in

Faeces ( % w/w) 2.7 3.8 3.3 3.3 3.2 3.0 2.8

P SD* 0.3 0.4 0.4 0.1 0.3 0.3 0.2

Ca:P Ratio in Faeces 1.4:1 1.2:1 1.1:1 1.1:1 1.2:1 1.3:1 1.3:1

Ca:P Ratio in Diet 2.4:1 1.6:1 1.6:1 1.5:1 1.7:1 1.6:1 1.7:1

* The Standard deviation is based on the difference between the four runs (8 for basal and 4 for the

other diets), however each run result was also analysed in triplicate and the average used for the

calculation.

Table 9 shows the Ca to P ratio in faeces and compares it to the diet ratio. The ratio

indicate the Ca concentrations were reduced in faeces particularly for the basal diet;

however, Ca to P ratio seem to be maintained at between 1.1:1 and 1.4:1, which

possibly shows a connection between the two elements.

75

3.3.3 The Analysis of Ca in Urine

The average amount of Ca excreted in the pig urine average was 500 mg/L, with the

maximum excreted about 2000 mg/L (Figure 8). The minimum total Ca requirement

NRC (1998) is 0.60 % w/w in the diet, whereas the amount of Ca in the diet varied

from 0.71 % w/w to 0.89 % w/w clearly indicating an excess of Ca in the diet. The

amounts of both Ca and P are regulated by the renal hormone calcitriol (1,25-

dihydroxycholecalciferol, 1,25-(OH)2D3). The hormone increases the absorption of

both Ca and P by stimulating the active transport systems in the intestinal tract.

When P absorption is restricted and Ca is freely available, hypercalcemia can

develop from increased Ca absorption (Schroder et al. 1996). However Bohlke et al

(2005) stated that the addition of limestone did not affect the apparent ileal and

apparent total tract digestibility for Ca. The majority of the Ca in all the diets was

supplied from the limestone, where in the basal diet (1) it supplied 0.72 % out of the

0.80 % total. Limestone contributed between 0.34 % for Palphos (2) and 0.59 % for

Kynofos (6). The amount of Ca excreted in urine for the basal and the Kynofos diets

were at opposite ends of the concentration range, with approximately 880 mg/L and

80 mg/L respectively, indicating additional factors (such as digestible P) influence the

amount of Calcium absorbed. Ca in the faeces for both these diets (basal and

Kynofos) were at 3.9 % w/w. Brana et al (2006) found that Ca absorption did not

change with increasing concentrations of available P. In this thesis P appeared to be

the limiting nutrient and hence unaffected by Ca.

76

-1000

-500

0

500

1000

1500

2000

2500

1 2 3 4 5 6 7

Diet Type

Con

cent

ratio

n (m

g Ca

/L)

Figure 8. Total calcium concentration in pig urine for each diet type (1 to 7)

3.4 Pig body weight

As a pig’s body weight increases, its mineral requirements (% or amount/kg of diet),

for Ca and P decrease (NRC 1998). Nevertheless their food intake increases. The

NRC pig feed tables show a reduction in P requirements when the body weight

category increased from 20 - 50 kg to 50 - 80 kg. The change in weight category

also gives an estimated feed intake increase from 1.855 kg to 2.575 kg per day. The

total Ca requirement drops from 0.60 to 0.50 % of diet and the total P drops from

0.50 to 0.45 % of diet. In this research the pigs were fed a constant and measured

1.5 kg of feed per day. A provision to increase this amount in the 3rd and 4th section

of the trial was available, but not acted upon. By maintaining the feed level at a lower

weight it both ensures minimal or no refusal of food, and sustained the daily P

amount below recommended requirements.

77

Table 10. Pig numbering allocated against the pig identification.

Pig ID A B C D E F G H

PRTC ID 22/7 7/9 10/7 11/5 22/10 11/4 8/7 21/8

Table 10 shows the identification of the individual pigs as well as the ear marking

identifications given to them by the PRTC at Werribee.

35

40

45

50

55

60

65

70

75

80

27/10

/2004

3/11/2

004

10/11

/2004

17/11

/2004

24/11

/2004

1/12/2

004

8/12/2

004

15/12

/2004

Date of Weighing

Pig

Wei

ght (

Kg)

G

B

F

D

H

E

C

A

Figure 9. Pig weight trends

Eight male pigs with a starting weight of 41.6 ± 0.4 kg (Standard Error of the Mean,

SEM) were randomly allocated a test diet (Table 1). The pigs were given 1.5 kg/day

for four consecutive 14 day periods, the final 5 day period (days 9 to 14) consisting

of a total faeces collection. Apart from a period of weight loss for two of the pigs (A

and D) (Tables 3 and 10, Figure 9), between the 19th and the 24th of November, all

78

pigs consistently gained weight. The period of weight loss was the 9 days between

the feeding trials, and so had no effect on the project. At the end of the final

experimental period, the pigs weighed 72.9 ± 0.9 kg (SEM) (Figure 9). In an eight

week period the pigs almost doubled their weight, increasing almost 8 kg for each 14

day trial period. Average weight gain over the 56 days was 0.56 kg/day, which is

below the NRC (1998) value of 0.80 kg/day based on energy levels.

79

3.5 Calculation of P-Digestibility.

Table 11. P- Digestibility results for each pig and diet.

Run

Pig

ID

Die

t

Am

ount

of n

utrie

nt

with

in th

e m

ix (A

)

(g/k

g)

Tota

l Int

ake

(I) (g

)

Tota

l Out

put (

O) (

g)

Dig

test

(%)

Ave

rage

Bas

al w

ithin

Run

s (%

) (D

igte

st A

B)

Dig

add (%

)

1 B 1 3.13 21.7 17.4 19.8

1 G 1 3.13 22.6 18.4 18.7 19.2

1 A 2 4.78 35.2 25.0 29.0 47.6

1 H 3 4.59 33.6 25.0 25.4 38.5

1 D 4 4.76 34.8 22.7 34.8 64.8

1 E 5 4.83 35.5 22.8 35.7 66.1

1 C 6 4.82 35.9 22.4 37.5 71.3

1 F 7 5.30 39.6 23.0 41.8 74.2

2 C 1 3.13 23.3 14.2 38.8

2 H 1 3.13 23.1 17.2 25.8 32.3

2 D 2 4.78 35.2 24.9 29.1 23.1

2 F 3 4.59 34.3 17.7 48.3 82.4

2 G 4 4.76 35.3 17.8 49.7 82.9

2 B 5 4.83 35.9 18.6 48.1 77.2

2 A 6 4.82 36.0 19.6 45.5 69.7

2 E 7 5.30 39.2 17.8 54.6 86.5

80

Table 11. Continued Run

Pig ID

Diet

Amount of nutrient

within the mix (A)

(g/kg)

Total Intake (I) (g)

Total Output (O) (g)

Digtest (%)

Average Basal within

Runs (%) (Digtest AB)

Digadd (%)

3 A 1 3.13 23.3 21.7 6.5

3 D 1 3.13 23.1 15.9 31.2 18.9

3 G 2 4.78 35.0 24.7 29.4 49.4

3 B 3 4.59 33.8 28.5 15.6 8.6

3 H 4 4.76 35.2 23.0 34.7 65.1

3 E 5 4.83 35.8 23.3 34.7 63.9

3 F 6 4.82 36.0 23.6 34.4 63.0

3 C 7 5.30 39.3 22.1 43.8 79.5

4 E 1 3.13 23.1 18.8 18.6

4 F 1 3.13 23.3 17.6 24.2 21.4

4 B 2 4.78 35.0 22.4 35.9 63.5

4 C 3 4.59 34.2 22.1 35.4 65.3

4 A 4 4.76 35.3 21.1 40.3 76.5

4 G 5 4.83 35.4 19.8 44.0 85.6

4 H 6 4.82 35.9 20.8 42.2 80.5

4 D 7 5.30 39.2 18.1 53.8 100.2

* The nutrient in this table is Phosphorus

Table 11 shows the data obtained for the four runs, and also diets 1 to 7 for each of

the pigs (A to H). Table 11 includes the amount of nutrient (Total P) in the mix (A) in

81

%, the total intake (I) (Equation 1) for each pig and the total output (O) (Equation 2).

The digestibility of the test diet, Digtest (Equation 3) was calculated from the two

previous results. The average basal result (Digtest AB) was then calculated and was

used to calculate the digestibility of the nutrient / P in the additive (Digadd ) (Equation

4). The total intake (I) was consistent between all 4 runs. The total output was lower

in run 2 than the other three runs for all but diet 2, (Palphos). Hence by using the

mean for the Basal diet in each run, results can be corrected for any conditions which

may have caused a difference between the runs. The output for the Palphos was

consistent for all 4 runs. Diet 3, the Rediphos DCP, had a very large standard

deviation. This was caused by a possible outlier in run 3, as this result caused a

significant reduction in the overall digestibility of this additive. This will be discussed

further, later in this section.

Table 12. Average basal digestible P result for each run, and t-test for run 2.

Run 1

(%)

Run 2

(%)

Run 3

(%)

Run 4

(%)

Mean

(%) SD* t-test t 0.05,3

19.23 32.30 18.87 21.42 23 6 2.95 2.353

*SD is calculated from the 4 results in the runs, and not the original 8 result.

The Basal diet results (Table 12), for the digestible P in the test diet (in percent) was

averaged for each of the 4 runs and used to calculate the digestible P for the test

ingredient for each run. It should be noted that the digestible P for the basal diets, in

the second run averaged at 32 percent, whereas the other 3 runs ranged between 19

and 21 percent (Table 11 and 12). The t-test indicated there was a significant

difference between run 2 and runs 1, 3 and 4. Consequently each run was then

treated as a separate set of results. By using the average basal result for each run

82

(instead of the overall mean of 23%) to calculate the P digestibility the standard

deviation was reduced significantly for most of the additives, as well as giving a better

ANOVA P-value from 0.17 to 0.025 and the P digestibility result remained the same.

Changing the calculation of the digestible P from the average of the total experiment

to four individual calculations based on each run, is likely to have reduced any

differences due to weight and age, as well as any environmental factors.

Table 13. Dietary P digestibility, means and standard deviations.

Diet Name Diet Number Mean (%) SD (%)

Basal 1 23.0 9.6

Palphos TCP 2 30.9 3.4

Rediphos DCP 3 31 14

Honge DCP 4 39.9 7.0

Bolifor DCP 5 40.7 6.5

Kynofos MDCP 6 39.9 4.9

QAF MBM 7 48.5 6.6

Table 13 shows the average P digestibility of each diet, prior to calculating the P

digestibility of the individual P additive. This was calculated using Equation 3. This

was then used to calculate the digestibility by difference, using Equation 4, the

results of which are shown in Table 14.

83

Table 14. P digestibility means and standard deviations.

Additive Mean (%) SD 1 (%)* SD 2 (%)*

Palphos TCP 46 17 10

Rediphos DCP

Run 3 result removed

49

65

32

22

44

36

Honge DCP 72 9 20

Bolifor DCP 73 10 18

Kynofos MDCP 71 7 14

QAF MBM 85 11 16

* The standard deviations are distinguished by the way the results had been determined, SD1 treats

each run as an individuals group, SD2 treats all results as one group by averaging the basal result .

Table 14 shows the mean digestibility for each additive, as well as two different

standard deviations of the population. These are distinguished by the way the final

results have been calculated; that is, the use of the overall mean of the basal result

for SD 2, and the mean of the basal result for each run, SD 1. To be more precise the

SD 2 treats all the results as one group, whereas SD 1 separates each run as

individual groups. The mean results for either of the two methods are identical. As

can be seen in Table 14, there is a significantly larger standard deviation for the

overall basal mean calculation than the mean of the separated runs. It was also

noted that only the TCP result gave a higher standard deviation using each run. It

can also be seen in Table 14, that the removal of the outlying result in run 3 for

Rediphos DCP will bring the mean closer to the other DCP results and significantly

reduces the standard deviation; however, as there is no known reason to remove this

result, it was not removed for future calculations. A t-test of run 3 for Rediphos gave a

result of 2.23 which is less than the t 0.05,3 of 2.353, however the t-test for run 2 for

Rediphos give 2.44 which is above the t 0.05,3, indicating that this result could be

84

removed. The Grubbs’ test at a level of significance set at 0.05 (two –sided) detected

no outliers, with the critical Z value of 1.48, run 2 had the furthest result with a Z

value of 1.22. On the basis of these findings, it would be better in future experiments

to increase the number of tests per diet, to improve the calculation of the mean for

any samples such as the Rediphos which may give unusually high variation.

The mean results in Table 14 give the apparent digestibility of P from the major

inorganic sources of P which are significantly less than 100 %. The digestibility of P

in the basal diet, a corn and soybean meal (Table 12), was 23 % (SEM = 4.3 %).This

was higher than had been quoted by the NRC (1998), where the digestibility of corn

is 14 % and the digestibility of soybean meal is 23%, so the result would expect to be

between these two result. The NRC (1998) results are Bioavailability of Phosphorus

(%) relative to mono-sodium or mono-calcium phosphate. The MDCP (Kynofos) used

was 71 % hence a relative result to this will be approximately 16%. Both the quantity

of total P and the availability in the corn will vary depending the variety, growing

region and other environmental factors, such as growing season and storage (Kim

2003; Kim et al. 2005). There were significant differences in the digestibility of P in

the various sources with phosphorus digestibility in the inorganic P sources ranging

from 46 to 73 %, while the phosphorus digestibility in meat and bone meal was 85 %

(Table 14). The apparent digestibility of P in rock phosphate was significantly less

than in most of the other inorganic sources of P. Frenseuie (2004) quoted 49 %

apparent absorption for tricalcium phosphate, and 39 % for sedimentary phosphate.

The apparent digestibility of P in most of the mono- and dicalcium phosphates was

similar and within the range Jongbloed (1993) reported for a wide range of feed

phosphates (65 % to 90 %, DCP at 65 %, DCP dihydrate at 69 %). Furthermore,

Jongbloed (1993) concluded that the apparent phosphorus digestibility in meat and

85

bone meal samples ranged from 74 to 85 % which is similar to the results of the

present study.

Kemme et al. (1994) reported that the apparent digestibility of P with pigs fed Bolifor

monosodium phosphate, dicalcium phosphate dihydrate and monocalcium

phosphate monohydrate was 87.3, 70.7 and 84.2 %, respectively. The lower

phosphorus digestibility value in Bolifor dicalcium phosphate dihydrate reported by

Kemme et al. (1994) is similar to the value found for the same product in this study.

Our product is also likely to be the dihydrate form due to the DM value resulting in a

hydration amount of 14.5 % (whereas a true dihydrate for DCP is about 17 % water).

Poulsen (1995) conducted digestibility studies to establish the apparent digestibility

of seven feed phosphates and meat and bone meal and her results revealed that the

digestibility of P differed among the 8 sources with phosphorus digestibility in Bolifor

dicalcium phosphate dihydrate being only 52 % compared to 73 % for the same

product in the current study, albeit 10 years later. The apparent phosphorus

digestibility found by Poulsen (1995) in meat and bone meal was also lower at 54 %.

Frenseuie (2004) quoted three different values, apparent absorption, apparent net

retention and utilization efficiency, of which the utilization efficiency used

monocalcium phosphate as a reference control at 100 %. Many other sources also

use a reference control, usually sodium phosphate or ammonium phosphate at 100

%, to calculate the digestible P. As their seems to be a number of different

techniques to measure digestibility, it can be easy for researchers to confuse or

compare different forms of digestibility. And if a reference control is used, how does

it compare to other samples used as reference controls? Can researches agree to

use a single high purity sample as a control?

86

The apparent phosphorus digestibility of the six sources employed in this research

were consistent with other similar studies(Jongbloed et al. 1993, Kemme et al. 1994,

Frenseuie et al. 2004), and clearly demonstrated that they are not 100 % digestible.

One of the DCP (Rediphos) gave a highly variable result which was lower than the

other DCP samples. This indicated differences between the Rediphos and the other

DCP samples. By treating each run as an individual experiment, the standard

deviation was reduced allowing improved comparative work against other techniques,

such as the in vitro methods in this study

3.6 The in vitro wet chemistry tests

The in vitro research involved comparing a range of selected common extractants

and then analysis of the results to observe if there are correlations between the wet

chemical analysis and the results of the in vivo experiments.

The in vitro wet chemical methods have been compared against the in vivo result for

the apparent digestibility of P. The chemical methods have been used to determine

the availability of P from P sources. The digestibility has been determined by the

percentage of P as phosphate dissolved in the solution. All samples were analysed

in groups of seven for each method. The results in Figure 10 (Appendix F showing

the mean and standard deviation) are indicative of the amount of sample or

ingredient which remained in the solution used for testing. The water and citric acid

extractions are regarded as equivalent to the P digestibility by the feed industry. The

results are also only indicative of the percentage of reactive phosphate P which may

not be the total P in the solution.

87

0

10

20

30

40

50

60

70

80

90

100

Rediph

os

Hongh

e

Bolifor

Palpho

ss

QAF-MBM

Kynop

hos

Sample types

% P

Dig

esta

bilit

y% PDigestability

CaCl2 1Hr 1:10

NaHCO3 1/2Hr1:20

NaHCO3 16Hr1:10

AmmAc 1/2Hr1:20Water 1g to250mL

2%Citric acid 1gto 200mLHCl Bicarb Mix0.25 g to 45 mL

HCl Bicarb Mix0.5 g to 45 mL

HCl Bicarb Mix0.75 g to 45 mLHCl Bicarb Mix1 g to 40 mL

Figure 10. Comparison of the in vitro tests for P digestibility in the six P

sources

The in vitro phosphorus digestibility for the various chemical methods (Figure 10 and

Appendix F), clearly show a difference between each of the DCPs with the MDCP

(Kynophos) being the most soluble in all extractants used. Nine additional sources of

P were also tested via the in vitro techniques (Appendix F), however they were

unable to be analysed via the in vivo technique. One sample of additional DCP from

Biofos was comparable to the Kynophos and Quantum MDCPs, but the Redox

MDCP performed as if it contained more of a less soluble form of DCP. The two

different sodium bicarbonate solutions showed that extraction time did make a

difference to the solubility for some of the P sources, the Palphos TCP and the QAF-

88

MBM, and Biofos DCP and Quantum MDCP all had a higher result even though the

sample solution ratio was smaller by half, indicating time as a factor (figure 11). The

results showed that the longer extraction time gave a reduced standard deviation, but

due to the high alkalinity the calcium phosphate remained relatively insoluble. As

found with the bicarbonate extraction, some forms of calcium phosphate are relatively

insoluble in alkaline solution. The two sodium bicarbonate solutions also had

different sample mass to extractant volume ratio, the effect of this can be seen later

in figure 12.

0102030405060708090

100

Redipho

s

Honghe

Bolifor

Palpho

ss

QAF-MBM

Kynoph

os

Sample types

% P

Dig

esta

bilit

y

% P Digestability

NaHCO3 1/2Hr 1:20

NaHCO3 16Hr 1:10

AmmAc 1/2Hr 1:20

2%Citric acid 1g to200mL

Figure 11. Comparison of extraction times (1/2 Hr and 16 Hrs) and pH

The pH of NaHCO3 = 8.0, of Ammonium Acetate = 7.0 and of 2 % Citric acid = 2.1.

The calcium phosphate will become soluble in acid solutions. This is shown in figures

10 and 11 where the 2 % citric acid solution with only the Palphos TCP being poorly

solubilised as it gave a result of 10 % whereas the Multifos TCP reached 84 %. The

citric acid method only discriminated between the rock phosphate sample and the

89

meat and bone meal, as the MDCP and DCP from the original set of samples used in

the in vivo experiment all gave results of approximately 97 % extracted P. In contrast,

some of the additional MDCPs and DCPs were much lower. This indicated

differences between the forms of DCP, or the inclusion of TCP or rock phosphate in

those samples. Therefore the citric acid method can be used to differentiate the

possible addition of TCP into the samples, as some DCP (as claimed by the

company) can be a mixture of MCP and TCP.

The acetate was expected to have a similar effect to the citrate in the citric acid (as

they will complex with the Ca), with the difference between the two methods being

the pH of the ammonium acetate (7.0) (figure 11) and a smaller sample to solution

ratio. As can be seen below in figure 12 the sample mass to extractant volume ratio

will not have such a significant effect and can be discounted as an effect in this

situation. The main reason that the samples in the acetate solution were only

sparingly soluble was due to the neutral or non- acidic nature of the solution. The pH

of the extractant is significant, whereas the complexing agent (citrate or acetate) is of

little significant in dissolving the DCP or TCP. Acetate is regarded as a fairly weak

organic complexing agent that will form ionic or electrostatic bonds with Ca. Other

methods from the feed industry which use citrate, include Petermann’s method (an

alkaline ammonium citrate), and an ammonium citrate at pH adjusted to 7.0. Figure

11 shows the effect of pH with the alkaline bicarbonate solutions and the neutral

acetate solution both having similar effects with solubility, whereby the acidic citric

acid solution solubilises the samples more easily.

90

0

20

40

60

80

100

Redipho

s

Honghe

Bolifor

Palpho

ss

QAF-MBM

Kynoph

os Sample types

%P

dige

stib

ility

% P Digestability

HCl Bicarb Mix0.25 g to 45 mL

HCl Bicarb Mix0.5 g to 45 mL

HCl Bicarb Mix0.75 g to 45 mL

HCl Bicarb Mix 1g to 40 mL

Figure 12. Comparison of sample mass to extractant volume ratio.

Figure 12 indicates the larger the sample mass to extractant volume ratio the more

soluble some of the samples become. This can also be seen in figure 13 where the

water extraction has a ratio of 1 : 250 compared to the calcium chloride extraction

which is only 1 : 10 although other effects may occur. Figure 11 containing two

different bicarbonate extractions, with different times as well as volumes also

suggested the higher volume ratio will contribute to solubility.

The digestive tract of the animals begins with an acid solution in the stomach, which

increases the solubility of the phosphate samples in the acid matrix. The feed then

passes into the small intestine where the addition of the bile solution makes it

alkaline. This point in the ileum is where P is absorbed. Bile digestive fluid is mainly

composed of carbonate salts. It was shown that the carbonate solution by itself

achieved minimal solubility. By mimicking the digestive system and thus taking

advantage of an acid matrix to first solubilise the phosphate, a greater amount of

91

phosphate will remain in solution when the pH is adjusted to 8.0 with bicarbonate.

Increased solubility was achieved by modifying the sample to solution ratio, except

for the Monofos MAP (Appendix F) which levelled out at about 81 %. Additional

improvement may also be achieved by adjusting extraction times, as in the

bicarbonate section of the test.

0102030405060708090

100

Redipho

s

Honghe

Bolifor

Palpho

ss

QAF-MBM

Kynoph

osSample types

% P

Dig

esta

bilit

y

% P Digestability

CaCl2 1Hr 1:10

Water 1g to250mL

Figure 13. Comparison of water and calcium chloride solution.

Calcium chloride is commonly used to extract for P from soil. Unfortunately, by

adding more Ca to calcium phosphate in the form of calcium chloride, the equilibrium

is being forced back to the insoluble calcium phosphate (in non acidic solutions), as

can be seen if we compare this extract with the water extract in figure 13 . This is not

the case with the MDCP samples (and Biofos DCP which acted like a MDCP sample)

which due to their acid nature assists in the solubility.

92

Table 15. Correlation matrix between in vitro and in vivo P digestibility of six

phosphorus sources

In v

ivo

P D

iges

t

CaC

l 2

Na

HC

O3 ½

hr

Na

HC

O3

16 h

r

NH

4 ace

tate

Wat

er s

ol.

2% c

itric

aci

d

HC

l / B

icar

b 1

g

HC

l / B

icar

b 0.

25 g

Correlation coefficient+

In vivo P digest 1

CaCl2 0.192 1

Na HCO3 ½ hr 0.190 0.996 1

Na HCO3 16 hr 0.226 0.996 0.998 1

NH4 acetate 0.217 0.994 0.999 1 1

Water sol. 0.188 0.998 1 0.998 0.998 1

2% citric acid 0.525 0.309 0.364 0.356 0.372 0.355 1

HCl/Bicarb 1.0 g 0.350 0.945 0.961 0.971 0.971 0.956 0.444 1

HCl/Bicarb 0.25 g 0.537 0.821 0.851 0.867 0.870 0.842 0.601 0.957 1

+ The correlation coefficient (R) needs to be ≥ 0.975 (for 95 % confidence or 2 standard deviations, or

R ≥ 0.816 for 67 % confidence of 1 standard deviation) for the correlations between in vivo P

digestibility and in vitro methods to be significant (P<0.05). Results with significant correlations have

been highlighted by italics.

While there were significant differences in the in vitro phosphorus digestibility

between the various chemical methods (Figure 10), the majority of the chemical

methods were not discriminative enough as illustrated by the very low and non-

significant correlation coefficients between the in vitro methods and in vivo

phosphorus digestibility (Table 15). The correlation coefficient between the citric acid

and the 0.25 g HCl / bicarbonate method were highest at 0.53 and 0.54 respectively

but well below the significance level (Table 15). The similarities between these two

93

methods are the high dilution factor and the acid matrix. The dilution factor alone will

not make much of a difference as the water soluble method had a higher dilution

factor (1:250) than all the methods, yet only had a correlation coefficient of 0.19. It did

however show a high solubility for the Kynophos (MDCP), a mixture of acidic mono

calcium phosphate and dicalcium phosphate. Thus the acid extraction must play an

important role in determining digestibility in the samples.

Dellaert et al. (1990) reported a consistent and high solubility (93 – 100 %) of P in 2

% citric acid while the solubility in water ranged from 1 to 100 %. The solubilities of

these two methods were unrelated to in vivo phosphorus digestibility for those P

sources, which included dicalcium, monocalcium, disodium and calcium sodium

phosphates. Kemme et al. (1994) also determined the solubility of the three Bolifor

products in water and citric acid and found that there was no relationship at all

between solubility and apparent phosphorus digestibility. Of the other chemical

methods investigated, the HCl / bicarbonate extraction provided some potential (R2 =

0.35 at 1.0 g). Further refinements of this method by increasing the dilution factor

showed only minor improvement in the correlation (R2 = 0.54 at 0.25 g) with in vivo P

data across the seven P sources (Tables 15 and 16). Further modification of this

method by the addition of chelating agent such as EDTA into the acid, and optimising

the sample to solution ratio as well as adjusting the extraction times may be of

benefit. Thus it is unlikely that the chemical methods tested can be used to predict

the phosphorus availability in a wide variety of P sources.

However, the in vitro chemical methods are (except the citric acid method and the

0.25g HCl / bicarbonate extraction) closely correlated to each other and seemed to

be able to differentiate some types of inorganic P sources (Table 15). The single

94

sample of mono-ammonium phosphate (MAP) had a very high solubility in all of the

in vitro P digestibility methods. The MAP did not have the expected increase in

solubility with increasing dilutions indicating it had reached its maximum solubility in

that particular matrix (HCl / bicarbonate). The Multifos TCP was more soluble than

the rock phosphate (Palphos TCP) especially for the citric acid extraction. The Biofos

DCP was more closely related to the results of the MDCP, and the Biophos fish meal

mixed with TCP was less soluble than the MBM. These differences in the chemical in

vitro methods would also suggest differences in the in vivo method; however the

correlation shows significant difference between the method types (in vivo versus in

vitro).

3.7 The Infrared analysis

Infrared (IR) analysis was selected as it is already commonly used in the feed

industry. The samples analysed by the in vivo experiments, were used to calibrate

the IR spectrometers and a self prediction test used to validate the results.

This research could only conduct a preliminary analysis of the potential usefulness of

NIR and/or MIR to predict in vivo P digestibility because only six P sources were

available for calibration and these techniques require large numbers of samples to be

used as calibrants.

NIR did not correlate with any of the in vivo results. Although correlation curves and

factor estimate curves (similar to those shown in Appendix H) for both uncorrected

data and the 1st derivatives of the data gave a correlation coefficient (R2) greater than

0.999, the predictions were not obtainable due to the large Mahalanobis Distance

95

(Thermo-Galactic, 2003) for all the “unknowns”, making them unrecognizable as a

match (for sample type) by the software. This distance is an indication of “fit” of the

existing data used for the calibration and such a large distance indicates that the

sample size was too small. The cross-validation self-prediction test, involves the

removal of an “unknown” sample (or samples) from the calibration set and then

predicting the result of this sample.

Figure 14 show examples of the MIR spectra for Rediphos DCP and Honghe DCP.

Appendix G shows MIR spectrum for Kynophos MDCP, Bolifor DCP, Palphos TCP

and QAF-MBM. It also includes other sources of P not used in the digestibility

experiment: Biophos (fish meal and TCP), Ridley DCP, Biofos MDCP and Quantum

MDCP.

7800.0 7000 6000 5000 4000 3000 2000 1500 1000 450.00.44

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.78

cm-1

A

Figure 14. MIR spectrum of Rediphos DCP (Black) and Honghe DCP (Blue)

* In the Y-axis the A is for Absorbance and the X-axis the units is wavenumber cm-1

The spectra suggest clear differences between the various types of compounds

96

MDCP, DCP TCP, and MBM. They also show large similarities among the 3 types of

MDCP tested, and all 4 DCP compounds show differences. Figure 14 shows the

MIR spectrum of two DCP samples which are clearly different, with Honghe DCP had

extra peaks at approximately 1300 cm-1 and 3400 cm-1. The Ridleys DCP had weak

absorbance compared to the other 3. It also showed closer similarities to the TCP

but having broader peaks. The other 3 (as well as the MDCP) had high absorbance

from approximately 3500 cm-1down to 450 cm-1. The PO43- ion is expected to absorb

strongly at 1100 – 1000 cm-1. All spectra showed peaks in this area (Appendix G).

The P=O stretching frequency has two peaks expected at about 1295 and 1285 cm-1

and should be independent of conjugation effects. Both the Palphos and the QAF-

MBM showed peaks at the P=O stretching frequency, Honge DCP had a small peak

around the 1280 cm-1, all other samples tested did not show signs of these peaks,

although the peaks may have been masked by other factors.

Due to the effect of using multiple spectra on the correlation coefficient (improving its

relationship with the curve), the two MIR spectra with least visible difference from

each in vivo sample were selected to produce an estimation of the curve. The

calibration curve/model with the number of factors set at 4, was determined using

prediction residual error sum of squares (PRESS) (refer to Appendix H) for all three

parameters (P digestibility, total P and total Ca). The factors were determined by

using partial least squares (PLS1). The calibration curves were then used reliably to

self predict the digestibility (via the cross validation with one spectrum removed)

(results for self prediction in Table 16 allows for comparison between the measured

and predicted results of the in vivo digestibility, results for self prediction were

obtained from both the unprocessed (straight) spectral data and the 1st derivative of

the spectra). The calibrations were then used to predict the digestibility for other

97

unknown samples (samples not used in the calibration and obtained from different

sources). Although these predictions appear to be reliable for the sample

digestibility, they will need to be verified by future experiments.

The results selected in Table 16 show the two sets of MIR data curves which gave

the best fit. The estimation/prediction curve using the unprocessed spectra gave a

correlation coefficient (R2) of 0.993, the 1st Derivative curve (a ‘simplified’ curve

calculated from the original curve in the spectra) gave a R2 of 0.978, both results

indicating the curves have very good fits (relationships between the points in the

curves). The data has also been analysed using the Student t-Test, where results

below 2.45 are equivalent to having a probability of less than 0.05 (P < 0.05),

indicating the two methods are not significantly different. There are some results in

the 1st derivative curve that have a t-test result above 2.45, however the variations in

the predictions show no practical significance due to the poor precision shown in the

in vivo results.

98

Table 16. Variation (mean, standard deviation) in in vivo P digestibility

measured by MIR in all phosphorus samples

In vivo P

digestibility (%)

Mean1 P digestibility (%) predicted by

Mid Infrared spectroscopy Phosphorus

Source

mea

n

S.D

Stra

ight

Spe

ctra

t-Tes

t2

Firs

t

deriv

ativ

e

t-Tes

t2

Rediphos DCP 48.7 27.9 48.4 0.99 55.4 3.61

Honghe DCP 72.3 7.7 72.5 0.17 67.3 2.99

Kynophos 71.1 6.2 71.6 0.68 70.0 1.26

Bolifor DCP 73.2 8.8 73.7 1.11 71.9 3.87

Palphos 45.9 14.5 46.9 1.37 47.1 1.47

QAF MBM 85.1 9.8 85.3 0.07 84.3 1.54

AFP DCP 49.0 70.6

Biofos MDCP 75.4 68.5

Biophos (Fish) 43.5 71.9

Monofos MAP 109.8 72.5

Multiphos TCP 22.6 56.6

Quantum DCP 6.1 58.6

Quantum MDCP 59.8 69.3

Redox DCP 48.8 58.9

Redox MDCP 72.2 70.3

PRTC MBM 83.5 83.9 1 Mean of spectra from a minimum of seven analogous spectra that were collected from each sample.

2 Student t-Test, where results below 2.45 not significantly different. The t-Test residuals indicate how

well the predicted concentration fits with the other samples. Results which were significantly different

have been highlighted by italics.

All the spectra (including those used in the calibration) were then analysed using the

calculated curves. There was close agreement between in vivo P digestibility and

99

mean P digestibility predicted by MIR (Table 16). Unfortunately, with the small

number of ingredients used in the pig trial, the prediction of digestibility could not be

performed on other samples, and can only be used as an estimate until more data

can be obtained to enhance the predictions. This is more evident when comparing

the results of the additional samples by the two predicted forms (Straight or

unprocessed spectra and the first derivative). The differences in P digestibility

between P sources were reduced when the first derivative was used. However the

results from the first derivative when comparing against the in vivo results had a

greater difference, and was confirmed by the t-test. In vivo digestibility data for the

ten other P sources shown in Table 16 are not available but examination of the mean

P digestibility predicted by MIR indicated that Monofos MAP would have the highest

P digestibility. Mono-ammonium phosphate is usually considered the standard

inorganic P source and NRC (1998) suggests 100% bioavailability of phosphorus in

mono-ammonium phosphate. However, as the MAP was not included in this study,

(and this brand did not have results from other studies) the in vivo digestibility cannot

be confirmed. In addition, the other meat and bone meal (PRTC MBM) has a MIR

predicted P digestibility of 83.5 % which is similar to the in vivo digestibility of P in

QAF MBM which was also highly digestible. Furthermore the MIR predicted P

digestibility of the TCP was poorly digestible and in the range of the in vivo

digestibility of P in the other rock phosphate, Palphos. Since Biophos is actually a

mixture of fish meal and TCP, it was expected to have results lying between the two

products; although the proportion of each product was unknown. Quantum DCP

showed an extremely low MIR digestibility result for a DCP, again without a proper

comparison this prediction cannot be confirmed.

The data in Table 16 tends to indicate the need for further study of the MIR curves.

100

This should preferably be done on samples containing P digestibility data to

determine more accurately which MIR curve (if not both or a combination of the two)

will give the best fit (thus giving more accurate predictions). Alternatively, more in

vivo studies could be performed (unfortunately further studies of this nature can no

longer be done at the PRTC in Werribee). The difference between the raw spectra

and the first derivative predictions appear to be significant, especially for some of the

extra samples. The first and second derivatives are used to remove baseline effects,

however depending on how the derivatives are produces they can also enhance

noise in the spectrum. The derivative spectra can make loading vectors difficult to

identify. Thus the derivative spectra are likely to have larger errors.

MIR spectroscopy does show more potential to estimate apparent P digestibility in P

sources than in vitro chemical tests, but it will require a larger number of sources and

digestibility results to calibrate.

101

Table 17. Total P and Total Ca measured by MIR in all phosphorus samples

Mean Total results predicted by

Mid Infrared spectroscopy Totals by ICP-ES

Straight Spectra First derivative Phosphorus

Source Total P

% w/w

Total Ca

% w/w

Total P

% w/w

Total Ca

% w/w

Total P

% w/w

Total Ca

% w/w

Rediphos DCP 18.3 24.5 18.4 24.6 17.6 23.4

Honghe DCP 18.7 28.8 18.3 27.7 18.2 26.7

Kynophos 21.1 16.8 21.1 16.6 20.3 16.8

Bolifor DCP 18.3 24.3 18.3 24.4 17.9 24.1

Palphos 17.2 38.6 17.0 38.1 17.0 38.0

QAF MBM 3.2 5.9 3.4 6.1 4.1 6.6

AFP DCP 20.2 24.8 17.3 25.6 18.2 21.6

Biofos MDCP 19.7 14.8 21.5 18.2 20.0 17.2

Biophos (Fish) 9.8 25.7 5.6 15.9 12.2 13.8

Monofos MAP 23.6 0.9 17.1 12.9 15.2 16.8

Multiphos TCP 17.7 29.9 16.4 27.3 17.4 27.0

Quantum DCP 18.4 30.3 16.3 28.0 13.9 25.3

Quantum MDCP 19.8 14.2 21.8 18.8 19.7 16.4

Redox DCP 18.3 23.9 17.6 25.8 17.8 22.7

Redox MDCP 20.9 19.4 19.4 20.9 18.8 18.7

PRTC MBM 3.0 6.4 3.0 6.9 3.5 6.2

102

MIR was also used to predict total phosphorus and total Ca (Table 17). The results

were compared against those obtained from nitric – perchloric acid digestion and

ICP-ES analysis used at DPI Werribee. This has confirmed the results obtained using

the prediction techniques for these two parameters (total Ca and total P).

The analysis of the results showed the self prediction (in vivo samples used in the

calibration) using the straight spectra has a correlation coefficient (R2) of 0.806, with

the largest difference being 0.4 % for P and 1.0 % for Ca. The first derivative was

more variable particularly for the P result in MBM (a different matrix) at 0.9 %, and

almost 2.0 % for the Ca value for the Honghe DCP, even though it had a better

correlation coefficient (R2) of 0.978. Bertrand (2002) reported a much lower

correlation coefficient (R2) of 0.60 for total P in soil, which may be due to the high

variability of soil types, and the variable nature of P associated in soils. The analysis

of the additional samples showed particularly large differences for the two matrices

which were not used in the calibration set, fish meal (+TCP) and MAP, although the

fish meal contained TCP and should have been similar to the MBM. If we remove

these two forms from the results all total P concentrations showed less than 2.9 %

difference using the straight spectra (or 14 % variation from the ICP-ES result). Total

Ca concentration showed a larger difference (4.6 %) (or 32 % variation from the ICP-

ES result).

The MAP predicted results were meaningless with calculated concentrations of 17 %

total P and 13 % total Ca; they should have been 24 % total P and 0.9 % total Ca

based on the ICP-ES results in Table 17. The Ca result was particularly erroneous

suggesting that if this type of testing is to be performed, samples in different matrices

should be included in the calibration set and then it can be used for predictions with

103

that particular matrix. The predictions were shown only to be reliable for the matrix

calibrated against (that is, they were only suitable for calcium phosphate).

The calibration obtained was used reliably to self predict for the digestibility, total P

and total Ca, and then to predict for other samples obtained from different sources,

and appeared to be reliable for the digestibility, but the total P and Ca were

unpredictable for non calcium phosphate sources, thus requiring more testing.

Overall, the results using the MIR could be used as a prediction tool to determine

digestibility, total P and total Ca, with the proviso that only the matrices used in the

calibration can be determined, and that the error in the result can have a large

variance.

3.8 Solid State 31P Magic Angle Spinning NMR

Solid state NMR was selected as it is a non- destructive method, on the other hand

liquid NMR requires very harsh extractants to dissolve and determine components in

the sample. NMR was also chosen due to its ability to differentiate between different

forms of P.

The analysis of 16 samples by Solid State 31P Magic Angle Spinning NMR, included

the 6 feed phosphate additives in the trial, 3 feed ingredients (Blood meal, Corn and

soybean meal), and 7 faeces samples incorporating the 6 phosphate feed diets and

the basal diet.

104

Table 18. Peaks and sidebands for 16 samples by SS 31P MAS NMR.

Sidebands (ppm) Sample Peaks

(ppm) +ve -ve

Blood meal -1.7 49.2 - 52.6

Corn 0.9 60.8 - 58.1

Soybean meal -0.3 57.4 - 58.1

Rediphos DCP 1.1, -1.4, -7.1 58.4, 60.9

118.0, 120.8

- 58.4, -61.1

-118.2, -121.1

Honge DCP -1.3 50.0, 101.4 - 52.9, -104.3

MBM 3.0

Kynofos MDCP 0.1, -4.4 43.7, 48.3

91.9, 96.3

-48.1, -52.7

-96.3, -100.7

Bolifor DCP

hydrated

1.3, -1.0 51.3, 54.6,

107.6

-51.7, -54.5,

-105.0

Faeces – Basal 1.2 29.4, 58.6 -27.2, -55.6

Faeces – Rediphos 1.2 38.5 -36.8

Faeces – Honge 1.1, -7.5 36.9, 72.1 -34.7, -70.1

Faeces – MBM 2.0 54.9 -51.4

Faeces – Palphos 3.0, -7.2 53.1 -48.5

Faeces – Kynofos 1.1, -7.6 44.4, 50.7 -48.6, -57.9

Faeces – Bolifor 1.1 36.3, 71.4 -34.2, -69.6

Blood meal had only 0.2 % P. The peak at approximately – 1.7 ppm was only about

four to five times that of the background noise, and most likely to be a mixture of

different phosphate groups, including inorganic P, ATP, nucleic acids, phospholipids

and other phosphorylated compounds, Sahyun (1933) suggested phosphoric esters.

105

The corn and the soybean meal spectra were very similar although the slightly wider

corn peak may indicate a double peak near 0.9 ppm. Both had two spinning

sidebands near + 61 and – 58 ppm. The corn and soybean meal P species are both

likely to be a form of phytate and hence the spectra are expected to be very similar

(Table 18). The fact that the chemical shifts are different indicates differences in the

phytate or its storage.

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Figure 15. SS 31P MAS NMR of Rediphos DCP

The P in feed additives all showed strong NMR peaks (Appendix H) due to high total

P (between 17 and 21 %). Except MBM with a total P of about 3 %, which should

have a similar spectrum to the blood meal with the major addition of hydroxyapatite,

whose peak is most likely to be 3.0 ppm? The MBM also had sidebands close to the

± 70 ppm but due to the poor resolution and high background noise they remain

106

undetermined. Rediphos DCP (Figure 15) actually had two large peaks at 1.1 and -

1.4 and a smaller peak at -7.1 ppm suggesting there are at least 3 or more different

types of phosphate in this mixture. The other spectra can be seen in Appendix H.

Honge DCP had only one large peak at -1.3 ppm. Kynofos MDCP had two distinct

peaks at 0.1 ppm and a slightly smaller one at -4.4 ppm, which is as expected as its

name suggests a mixture of both MCP and DCP. MCP being more acidic is most

likely to have the chemical shift at -4.4 ppm. The Bolifor hydrated DCP also two

peaks, the main at 1.3 ppm which is the same as one of the Rediphos peaks, and a

smaller one at -1.0 ppm. Altogether there were 6 different phosphate peak positions

(6 or more different compounds) for the mono and dicalcium phosphate sources

tested. Due to unexpected truncation of the data in FID (Free induction decay) at

under 0.002 seconds (the average FID lasted approximately 0.005 seconds), a

proper spectrum for Palphos TCP could not be obtained, although an unconfirmed

peak was detected around 9 ppm.

The faeces spectra peaks, although much smaller than those of the phosphate

additive spectra (due to the lower total P value of about 3 to 4 %) were still distinct.

There were no peaks in faeces corresponding to the same chemical shift (in ppm) as

those found in the additive. All the inorganic P additives that were digested by the

animal come out with different chemical shifts, indicating a high probability that

changes occurred to the sample in the digestive system. Even the positions of the

sidebands differed, with the sources approximately ± 50 ppm and faeces sidebands

all varied from about ± 30 to ± 50 for the Palphos and MBM only. This indicated a

transformation or altering of the compounds as the peaks positions shifted after

passing through the pigs (possible change in pH, or ion interaction).

107

Examination of the difference in the chemical shifts (Table 18 and Appendix I) of the

faeces compared to each other and of the undigested inorganic P source. The basal

or corn started with a peak at 0.9 ppm and ended with a peak at 1.2 ppm. The faeces

from Rediphos and Bolifor also ended with peaks at 1.2 ppm and 1.1 ppm

respectively, suggestion the possibility that the digestive process converted the P to

similar waste products. Hence the question is the shift at 1.1 for the Rediphos and

Bolifor a derivative of phytate? Although the sidebands on the basal faeces have

different chemical shifts to the other faeces. This was not the case as both the MBM

and Palphos had different chemical shifts of 2.0 ppm and 3.0 ppm + 7.2 ppm

respectively.

The use of the solid state NMR enabled the purity of different compounds of P to be

determined, and demonstrated that Rediphos actually produced 3 peaks and thus 3

different forms of P. The smaller peak at -7 ppm is most likely due to less soluble

aluminium phosphate. It also indicated a peak shift after the P compounds passed

through the pig. Thus the P compound may have been altered by the digestion

process, or the peak position is altered by something in the faeces.

108

Chapter 4

CONCLUSION

109

4.1 Conclusions

Overall the project objectives were achieved where a number of significant outcomes

that can provide important information to the pig industry and also agricultural

industries and analytical organisations. The main objective of this research was to

provide data for apparent digestibility of phosphorus from 6 major inorganic sources

of phosphorus to the Australian Pork industry. This was successfully undertaken

using in vivo approach in which digestibility was determined on the basis of diet, urine

and faecal analysis using standard experimental protocols.

The ingredients used in each diet were analysed for Ca and P to determine the

amount of each ingredient required for each diet. These results assisted in

confirming the type of compound used; Palphos was consistent with the results for

TCP, and the moisture determined by DM % indicated the Bolifor was a hydrated

DCP.

The analysis of the different inorganic sources of P revealed commonly accepted

values should be taken only as a crude guide. Indicating analyse of these sources

should be considered for a more accurate result. The animal feeds in the in vivo

studies were based on a corn and soybean meal supplemented with amino acids,

vitamins and minerals. The concentrations of calcium and phosphorus were chosen

for consistency and to obtain maximum absorption.

The dietary mixtures then had their concentrations of Ca and P confirmed, although

diet 7, the QAF-MBM contained higher than expected values. The DM % of the diets

averaged 87.4 %. These results are involved in the calculation of input (Equation 1

110

page 43).

The trial consisted of two basal diets and six other diets containing the major

inorganic sources. They were randomly selected and fed to 8 pigs over 4 sequential

runs of 14 days. The pig’s body weight measurement indicated that the pigs grew at

a healthy rate from approximately 40 kg to approximately 70 kg.

Urine analysis confirmed deficiency of P in all diets. The faeces gave higher

concentrations for Palphos for both Ca and P. These results the calculation of output

(Equation 2 page 44).

All the results were then combined to calculate the apparent P digestibility. The

apparent P digestibility was determined from the difference between the input (how

much P was eaten) minus the output (what came out in the faeces). It was

calculated that the basal diet showed a higher than expected apparent P digestibility

(23 %). The apparent P digestibility of the inorganic sources were rock phosphate 46

%, the MBM 85 %, MDCP 71 % and three DCPs ranged from 49 % to 73 %. It was

found that the errors (standard deviation) in the inorganic source results were

reduced by treating each run as a separate set of results, which was confirmed by

performing a t-test on the average basal results for each run.

One of the DCPs (Rediphos) gave a very large variance. This may have been due to

impurities as it had lower Ca and P values. The samples were not tested for

undesirable substances, such as fluorine, cadmium or arsenic, so it was not possible

to determine the true cause of the variance. Rediphos did however contain

approximately 1 % w/w of magnesium and sulphur.

111

Six different in vitro extraction techniques were developed, some with different

sample weight to extractand volume ratios. These were compared and assessed

against the in vivo technique in order to determine whether they could predict the

apparent digestibility of phosphorus from the 6 inorganic phosphorus sources. These

in vitro chemical methods were designed with the intention of providing an easier and

inexpensive alternative to the in vivo techniques. Based on this research there was

poor correlation between the two approaches. The in vitro techniques included two

commonly used feed extraction techniques using water solubility and extraction with

2 % citric acid. Additional techniques used calcium chloride, ammonium acetate and

sodium bicarbonate extractions, and a double extraction using hydrochloric acid

(HCl) followed by sodium bicarbonate to imitate the digestive tract. The R2 values for

the comparison with the apparent P digestibility were very low and not significant.

The 2% citric acid technique could be used to differentiate TCP from the other P

forms tested, and the water solubility technique could be used to determine MDP

levels as it is highly soluble (thus the inorganic P forms could be differentiated and

possibly predict P digestibility). The HCl / bicarbonate mixture extraction involving

0.25 g of sample could show promise with further modification. A similar technique

has been used for corn (Spencer et al. 2000, Liu et al.1997), and could be trialled on

different compounds such as the inorganic sources.

The alternative to the chemical methods are the non-destructive spectroscopic

techniques. NIR and MIR both showed promise with high R2 for their correlation

curves and factor prediction curves. However, the cross-validation technique for NIR

was poor, and self prediction was unable to obtain any values. MIR was able to self

predict, however, the additional samples tested did not have any apparent P

digestibility data to be compared against. Total Ca and P concentrations were able to

112

be predicted, and showed acceptable results for calcium phosphates, but not for

other forms of phosphate such as ammonium phosphate (MAP). These

spectroscopic methods appeared promising, but require further evaluation. Hence

this research was unable to recommend an efficient, accurate in vitro chemical

method with the potential to assess the availability of phosphorus in inorganic

phosphorus sources. Further research into the spectroscopic methods is

recommended, in particular the use of MIR spectra.

Solid state 31P magic angle spinning NMR indicated Rediphos contained 3 different

forms of P, the smaller peak possibly being aluminium phosphate, however, this will

require further testing to confirm this. NMR analysis of the faeces showed the P

peaks was in a different position to the peak from the original inorganic P source,

suggesting a possible change in the chemical form.

Overall it was concluded that the apparent digestibility of phosphorus for the six

different phosphorus sources was between 46 and 85 %, which is below 100 % as

assumed by some growers. The results will allow the feed mixes containing these

sources to be more accurately manufactured allowing closer production to the

recommended concentrations by the NRC (1998). The NRC (1998) reported

universities and feed companies would add P from 10 to 55 percent above its

recommendations. By knowing exactly what is in the mix, the safety margins used

can be lowered or removed, thus lowering the addition of P sources to the mix, inturn

reducing the amount of P in the effluent produced by the piggeries.

For Example if 1.00 kg of DCP (100 % available) per 100 kg of feed to produce 0.50

% P as per the NRC (1998) nutrient requirements for pig between 20 and 50 kg, a

113

safety margin of 50 percent will require 1.50 kg of DCP. If DCP was known to be 75

% available then removal of the safety margin would allow a lower mass of 1.33 kg,

thus reducing the amount of P in the faeces. If however the DCP was only 50 %

available then a greater amount of DCP required would be 2.00 kg thus adding more

P to the faeces. This assumption is based on DCP being the only source of P in the

fed, and hence the true representation will be a lot lower.

4.2 Future work

The in vitro chemical methods examined in this project may be able to differentiate

inorganic phosphorus sources but cannot be used to estimate apparent P digestibility

of phosphorus sources. The ability to differentiate the different inorganic phosphorus

sources may prove useful. The currently used methods of water (most samples

except the MDCP were sparingly soluble) or 2 % citric acid (of which all but the TCP

were almost completely soluble), can be used together to differentiate between MCP

(soluble in water, and citric acid), DCP (only soluble in citric acid), and TCP (insoluble

in both). Further modification of the HCl / bicarbonate method, perhaps with the

addition of a chelating agent, and possibly different extraction times may prove

useful, as changing only the extraction ratio improved the correlation coefficient from

0.35 to 0.54. Exploration of the methods used by Liu (1997 and 1998) and Spencer

(2000) using HCl and succinate, would also be beneficial.

NIR and MIR spectroscopy do show some potential to assess the availability of

phosphorus in various phosphorus sources. This project was able to provide the NIR

and MIR calibrations based on only six sources of phosphorus. Validation of these

114

calibrations against other sources of phosphorus with known apparent P digestibility

data may be useful to determine if NIR/MIR could be used to accurately predict in

vivo P digestibility in phosphorus sources for pigs. Samples of known in vivo P

digestibility could be collected from other laboratories, analysed by NIR and MIR and

their predictions validated against the spectral calibrations determined in this project

to determine whether MIR and/or NIR could be used to reasonably predict in vivo P

digestibility of phosphorus sources.

MAP predicted results gave inaccurate results of 17 % total phosphorus and 13 %

total calcium, where they should have been 24 % total phosphorus and 1 % total

calcium. The calcium result was particularly in error. It should be suggested that if

these samples are to be tested they be included in the calibration set and then they

can be used for predictions with that particular matrix. The predictions were shown

only to be reliable for the matrix calibrated against (that is, they were only suitable for

calcium phosphate).

This poses the question of whether a sample similar to a blank can be used for these

prediction tests, and how close to the matrices used in the calibration must the

samples be? Can MAP be used as a total calcium blank when compared to DCP?

Can limestone (calcium carbonate) be used as a blank for the total P and hence the

digestible P when calibrating with calcium phosphates?

Unfortunately this research was not able to explore these ideas. And with the

retirement of Dr Ray King and the expected closure of the PRTC this research is

unable to continue at Werribee.

115

4.3 Implications

If calcium and phosphorus concentrations of common meat and bone meal are

considerably less than the commonly accepted values of 100 g Ca/kg and 50 g P/kg,

formulated diets based on this source may be marginal in calcium and/or

phosphorus. As meat and bone meal is often a major contributor to calcium and

phosphorus requirements in Australian pig diets, nutritionists and pork producers

should regularly analyse meat and bone meal (as well as other inorganic

phosphorous sources) they use for Ca and P content.

4.4 Recommendations

Nutritionists and pork producers should accurately determine the total calcium and

total phosphorus concentrations in the meat and bone meal and other inorganic

phosphorous sources that they use in their pig diets.

If nutritionists and pork producers formulate their diets on an available phosphorus

basis, they should not assume 100% availability in all inorganic P sources. They may

use the approximations on apparent P digestibility in phosphorus sources in Table

19. However due to the variability of the DCP and that only one of the other sources

types (that is only one TCP, one MDCP, and one MBM) were tested it would be

recommended that test be carried out for any source that has not been determined

by in vivo testing.

116

Table 19. Suggested P digestibility (%) of a generic P source

Generic P source Suggested P digestibility (%)

Meat and bone meal 85

Dicalcium phosphate 45 to 75

Mono - dicalcium phosphate 75

Rock phosphate 45

Assumptions have been made in the chemical analysis for the in vitro comparisons

are in the form of hydrolysable phosphate as detected by the colourimetric analysis.

NMR may indicate this is not the case as there was more than one peak for some of

the sources, although they still may be a hydrolysable form. Determination of ‘total P’

instead of (or as well as) ‘reactive P’ may have been preferable. Techniques such as

ICP-ES analysis or the inclusion of a chemical oxidation/digestion step for the

solution will achieve ‘total P’. Other techniques that may be used would be X-ray

diffraction (XRD) or X-ray fluorescence (XRF) to quantify the mineralogy of the

samples.

Greater research should be performed on the forms of P in the faeces, in particular

with changes in the diets and the availability of these forms to the soil and their

solubility and potential to change, and also the effect they have on water quality.

117

REFERENCES:

Ajakaiye, A, Fan, MZ, Archbold, T, Hacker, RR, Forsberg, CW & Phillips, JP 2003,

Determination of true digestive utilization of phosphorus and the endogenous

phosphorus outputs associated with soybean meal for growing pigs. Journal of

Animal Science, Vol. 81 no. 11 pp.2766 - 2775

AOAC, 1995, Official methods of analysis of AOAC International, 16th ed, Arlington,

Va, AOAC International.

APL, 2004, National Environmental Guidelines for Piggeries, Australian Pork Limited,

viewed 07 December 2004 http://www.australianpork.com.au

ARC, 1981, The nutrient requirements of pigs, Technical Review Revised Edition

Agricultural Research Council, Slough, England, Commonwealth Agricultural

Bureaux, xxii, 307 pp.

ASoE, 2001, Australian State of the Environment, Independent Report to the

Commonwealth Minister for the Environment and Heritage, CSIRO Published on

behalf of the Department of the Environment and Heritage, Canberra.

AUSPIG, 2006, AusPig, Australian Pork Limited and Pork CRC viewed 17 August

2006, http://www.australianpork.com.au/pages/page100.asp

Bertrand, I, Janik, LJ, Holloway, RE, Amstrong, RD & McLaughlin, MJ 2002, The

rapid assessment of concentrations and solid phase associations of macro- and

118

micronutrients in alkaline soils by mid-infrared diffuse reflectance spectroscopy,

Australian Journal of Soil Research, vol. 40 pp. 1339 - 1356.

Boblke, RA, Thaler, RC & Stein, HH 2005, Calcium, phosphorus, and amino acid

digestibility in low-phytate corn, normal corn, and soybean meal by growing pigs,

Journal of Animal Science, vol. 83 pp. 2396 - 2403.

Bollinger, DW, Tsunoda, A, Ledoux, DR, Ellersieck, MR & Veum, TL 2004, A simple

in vitro test tube method for estimating the bioavailability of phosphorus in feed

ingredients for swine, Journal of Agricultural and Food Chemistry, vol. 52 pp. 1804 -

1809.

Brana, DV, Ellis, M, Castaneda, EO, Sands, JS & Baker, DH 2006, Effect of a novel

phytase on growth performance, bone ash, and mineral digestibility in nursery and

grower-finisher pigs, Journal of Animal Science, vol. 84 pp 1839 - 1849.

Cade-Menun, BJ 2005, Characterizing phosphorus in environmental and agricultural

samples by 31P nuclear magnetic resonance spectroscopy, Talanta, vol. 66 no. 2 pp.

359-371.

CEFIC, 1999, A Guide to Feed phosphates, Scientific Committee, Inorganic Feed

Phosphates Sector Group, European Chemical Industry Council (CEFIC).

Cho, G, Wu, Y & Ackerman, JL 2003, Detection of Hydroxyl Ions in Bone Mineral by

Solid-State NMR Spectroscopy, Science, vol. 300 (5622) pp. 1123-1127.

119

Colwell, JD 1963, The Estimation of the phosphorus fertilizer requirements of wheat

in southern New South Wales by soil analysis, Australian Journal of Experimental

Agriculture and Animal Husbandry, vol. 3 pp. 190 - 198.

Condron, LM, Frossard, E, Newman, RH, Tekely, P & Morel J-L 1997, Use of 31P

NMR in the study of soils and the environment, Nuclear Magnetic Resonance

Spectroscopy in Environmental Chemistry, Oxford University Press, New York.

Corbridge, 1978, Phosphorus : an outline of its chemistry, biochemistry and

technology, Amsterdam, New York, Elsevier Scientific Pub. Co., New York,

distributors for the U.S. and Canada, Elsevier-North-Holland.

Cromwell, GL & Coffey, RD 1991, 'Phosphorus - A Key Essential Nutrient, Yet a

Possible Major Pollutant - Its Central Role in Animal Nutrition'. Proceedings of

Alltech's Seventh Annual Symposium, Lexington, Kentucky, Alltech Technical

Publications.

De Groote, G, Lippens, M, Jongbloed, AW & Meschy, F 2002, Study on the

bioavailability of major and trace minerals. M. Vanbelle. Brussels, Emfema

(International Association of the European (EU) Manufacturers of Major, Trace and

Specific Feed Mineral Materials), pp. 1 - 108.

Dellaert, BM, Van Der Peet, GFV, Jongbloed, AW & Beers, SA 1990, A comparison

of different techniques to assess the biological availability of fed phosphate in pig

feeding, Netherlands Journal of Agricultural Science, vol. 38 pp. 555 - 566.

120

Dilger, RN & Adeola, O 2006, Estimation of true phosphorus digestibility and

endogenous phosphorus loss in growing pigs fed conventional and low- phytate

soybean meals, Journal of Animal Science, vol. 84 pp. 627 - 634.

Edwards, AC 1996, Project BMI 5P, DPI Werribee, Pig Research Training Centre.

EPA, U. S. 1979, Phosphorus, All Forms. Method 365.1 (Colorimetric, Automated,

Ascorbic Acid), pp.365-1.1 -- 365-1.7. Methods for chemical analysis of water and

wastes, U. S. Environmental Protection Agency.

Esbensen, KH 2004, Multivariate Data Analysis - In Practice 5th Edition An

Introduction to Multivaiate Data Analysis and Experimental Design, CAMO Software

AS, Oslo, Norway.

Frenseuie, D, Chicco, CF & Garmendia, SGY 2004, Phosporus sources in swine

feeding. 2. Digestibility, Revista Cientifica - Facultad de Ciencias Veterinarias, vol. 14

no. 3 pp. 201 - 206.

Fogg, DN & Wilkinson, NT 1958, The colorimetric determination of phosphorus,

Analyst, vol. 83 pp.406 - 414.

Fuller, MF 1991, in vitro digestion for pigs and poultry, Wallingford, UK,

Commonwealth Agricultural Bureaux, International.

121

Gibson, DM & Ullah, ABJ 1990, Phytases and Their Action on Phytic Acid, Plant

Biology Vol. 9 Inositol Metabolism in Plants. J. Morre, W. F. Boss and F. A. Loewus,

Wiley-Liss, A john Wiley & Sons, inc. vol. 9 pp. 77 - 92.

Haaland, DM & Thomas, VT 1988, Partial least-square methods for spectral analysis.

I. Relation to other quantitative calibration mentods and the extraction of quanlitative

information, Analytical Chemistry, vol. 60 pp. 1193-1202.

Halmann, M (ed.) 1972, Analytical chemistry of phosphorus compounds, Chemical

Analysis, vol. 37, New York, Wiley-Interscience .

Hastad, CW, Dritz, SS, Tokach, MD, Goodband, RD, Nelssen, JL, DeRouchey, JM,

Boyd, RD & Johnstonet, ME 2004. Phosphorus requirements of growing-finishing

pigs reared in a commercial environment, Journal of Animal Science, vol. 82 no.10

pp. 2945-2952.

Hatzack, F, Hubel, F, Zhang, W, Hansen, PE & Rasmussen, SK 2001, Inositol

phosphates from barley low-phytate grain mutants analysed by metal-dye detection

and HPLC and NMR, Journal of Biochemistry, vol. 354 pp. 473 - 480.

Heim, D & Kihm, U 2003, Risk management of transmissible spongiform

encephalothies in Europe, OIE Revue Scientifique et Technique, vol. 22 no. 1 pp.

179-199.

122

Hunger, S, Cho, H, Sims, JT & Sparks, DL 2004, Direct Speciation of Phosphorus in

Alum-Amended Poultry Litter: Solid-State 31P NMR Investigation, Environmental

Science and Technology, vol. 38 no.3 pp. 674-681.

Hunger, S, Sims, JT & Sparks, DL 2005, How Accurate is the assesssment of

phosphorus Pools in poultry litter by sequential extraction? Journal of Environmental

Quality, vol. 34 no. 1 pp. 382 - 389.

Jerschow, A, Laws, DD & Bitter, H_ML 2002, Solid-state NMR Spectroscopic

Methods in Chemistry, Angewandte Chemie Int. Ed. vol 41 pp. 3096 - 3129.

Jongbloed, AW 1987, 'Phosphorus in the feeding of pigs: effect of diet on the

absorption and retention of phosphorus by growing pigs', Agricultural University, The

Netherlands, Wageningen.

Jongbloed, AW, Everts, H & Kemme, PA 1993, Recent developments in Pig Nutrition

2, Loughborough, UK, Nottingham University Press.

Jongbloed, AW 1997, Digestility of phosphorus in pig nutrition - experiences in the

Netherlands, Kraftfutter, no. 7 - 8 pp. 319 - 324.

Jongbloed, AW & Lenis, NP 1998, Environmental Concerns About Animal Manure,

Journal of Animal Science, vol. 76 pp. 2641 - 2648.

Kamphues, J, Zentek, J, Oberthur, RC, Flachowsky, G & Coenen, M 2001, Animal-

derived feeds as possible vectors for bovine spongiform encepholothy (BSE) in

123

Germany. 1. Comparative risk assessment for a single animal food of animal origin,

Deutsche Tierarztliche Wochenschrift, vol. 108 no. 7 pp. 283-290.

Kemme, PA, Jongbloed, AW & Mroz, Z 1994, 'Apparent digestibility of phosphorus

with pigs of Bolifor MSP, Bolifor DCP and Bolifor MCP', Lelystad, The Netherlands,

Rapport IVVO no. 389 Instituut voor Veehouderij en Diergezondheid.

Kemme, PA, Lommen, A, De Jonge, LH, Van der Klis, JD, Jongbloed, AW, Mroz, Z &

Beynen, AC 1999, Quantification of inositol phosphates using P-31 nuclear magnetic

resonance spectroscopy in animal nutrition, Journal of Agricultural and Food

Chemistry vol. 47 no. 12 pp. 5116-5121.

Kerr, M & Rayner, C 1995, 'Determination of Blue-Green Algal Toxins in Potable

Water Supplies', 1995 Government Food Analysts Conference, State Chemistry

Laboratory Werribee, Victoria.

Kim, JC 2003, 'Factors influencing the total tract apparent digestibility of nutrients

from wheat for the weanling pig', Division of Health Science, School of Veterinary and

Biomedical Sciences, Murdoch University.

Kim, JC, Simmins, PH, Mullan, BP & Pluske, JR 2005, The effect of wheat

phosphorus content and supplemental enzymes on digestibility andgrowth

performance of weaner pigs, Animal Feed Science and technology, vol. 118 no.1-2

pp. 139-152.

124

Koopmans, GF, Chardon, WJ, Dolfing, J, Oenema, O, Van der Meer, P & Van

Riemsdijk, WH 2003, Wet Chemical and Phosphorus-31 Nuclear Magnetic

Resonance Analysis of Phosphorus Speciation in a Sandy Soil Recieving Long-Term

Fertilizer or Animal Manure Applications, Journal of Environmental Quality, vol. 32

pp. 287 - 295.

Kopinski, J 2002a, 'Phosphorus for pigs', DPI&F Note no. P0057, Department of

Primary Industries Queensland.

Kopinski, J 2002b, 'Minerals for pigs', DPI&F Note no. P0058, Department of Primary

Industries Queensland.

Leinweber, P, Haumaier, L & Zech, W 1997, Sequential extractions and 31P-NMR

spectroscopy of phosphorus forms in animal manures, whole soils and particle-size

separates from a densely populated livestock area in northwest Germany. Biology

and Fertility of Soils, vol. 25 no. 1 pp. 89-94.

Liu, J, Ledoux, DR & Veum, TL 1997, In Vitro Procedure for Predicting the Enzymatic

Dephosphorylation of Phytate in Corn-Soybean Meal diets for Growing Swine.

Journal of Agriculture and Food Chemistry, vol. 45 pp. 2612 - 2617.

Liu, J, Ledoux, DR & Veum, TL 1998, In Vitro Prediction of Phosphorus Availability in

Feed Ingredients for Swine. Journal of Agriculture and Food Chemistry, vol. 46 pp.

2678 - 1681.

125

McDowell, RW & Stewart, I 2005, Phosphorus in fresh and dry dung of grazing dairy

cattle, deer, and sheep: sequential fraction and phosphorus-31 Nuclear Magnetic

Resonance Analyses. Journal of Environmental Quality, vol. 34 no. 2 pp. 598 - 606.

MLA 2003, 'Nutritional technical review'. Australian Meat and Bone Meal, Meat and

Livestock Australia Ltd.

Molins, RA 1991, Phosphates in food. Boca Raton : CRC Press. Chemical Methods,

Australian Soil and Land Survey Handbooks ; vol. 3, Melbourne : Inkata Press.

Morgan, MF 1941, 'Bulletin No. 450', Connecticut Agricultural Experimental Station:

479.

Mulder, L & Jongbloed, AW 1985, 'Determination of the value of feed phosphate in

pig feeding : literature rewiew and an experiment with piglets'. Lelystad, The

Netherlands, Rapport IVVO no. 169 Instituut voor Veehouderij en Diergezondheid.

Nahm, KH 2004, Additives to reduce phosphorus excretion and phosphorus solubility

in poultry and swine manure. Australian Journal of Experimental Agriculture, vol. 44

pp. 717 - 728.

NRC 1998, Nutrient Requirements of swine. National Research Council Tenth

Revised Edition. National Academy Press, Washington, DC.

126

Olsen, SR, Wateanabe, KS & Dean, LA 1954, 'Estimation of available phosphorus in

soils by extraction with sodium bicarbonate'. US. Department of Agriculture, Circular

No. 939.

Ozanne, PG & Shaw, TC 1967, Phosphate sorption by soil as a measure of the

phosphate requirement fo pasture growth. Australian journal of Agricultural Research,

vol. 18 pp. 601-612.

Pawitan, Y, Griffin, JM & Collins, JD 2004, Analysis and prediction of the BSE

incidence in Ireland. Preventive Veterinary Medicine, vol. 62 pp. 267 - 283.

Pettey, LA, Cromwell, GL & Lindemann, MD 2006, Estimate of endogenous

phosphorus loss in growing and finishing pigs fed semi-purified diets. Journal of

Animal Science, vol. 84 no. 3 pp.618 - 626.

Potter, LM, Potchanakorn, M, Ravindran, V & Kornegay, ET 1995, Bioavailability of

phosphorus in various phosphate sources using body weight and toe ash as

response criteria. Poultry Science, vol. 74 pp. 813-820.

Poulsen, HD 1995, Phosphorus digestibility in phosphates and meat and bone meal.

Report No. 34, National Institute of Animal Science, Denmark,.

Prevolnik, M, Candek-Potokar, M, Skorjanc, D, Velikonja-Bolta, S, Skrlep, M,

Znidarsic, T & Babnik, D 2005, Predicting intramuscular fat content in pork and beef

by near infrared spectroscopy. Journal of Near Infrared Spectroscopy, vol. 13 pp. 77-

85.

127

Qiao, Y & Van Kempen, T 2004, Technical note: Comparison of Raman, mid, and

near infrared spectroscopy for predicting the amino acid content in animal meals.

Journal of Animal Science, vol. 82 pp. 2596- 2600.

Rayment, GE & Higginson, FR 1992, Australian Laboratory Handbook of Soil and

Water Chemical Methods. Melbourne, Inkata Press.

Reeves III, JB 2001, Near- Versus Mid-Infrared Diffuse Reflectance Spectroscopy for

Determination of Minerals in Dried Poultry Manure. Poultry Science, vol. 80 pp. 1437-

1443.

Sahyun, M 1933, Blood phosphate. Journal of Biological Chemistry, vol. 101 pp. 295-

299.

Sands, JS, Ragland, D, Baxter, C, Joern, BC, Sauber, TE & Adeola, O 2001,

Phosphorus bioavailability, growth performance, and nutrient balance in pigs fed high

available phosphorus corn and phytase. Journal of Animal Science, vol. 79 pp. 2134

- 2142.

SCA 1987, 'Feeding Standards for Australian Livestock: Pigs'. Standing Committee of

Agriculture. East Melbourne. Australia., CSIRO.

Schneider, BH & Flatt, WP 1975, The Evaluation of feeds through digestibility

experiments, The University of Georgia Press, Athens, USA. pp.121-129 & 140 - 143

128

Schroder, B, Breves, G & Rodehutscord, M 1996, Mechanisms of intestinal

phosphorus absorption and availability of dietary phosphorus in pigs. Deutsche

Tieraerztliche Wochenschrift, vol. 103 no. 6 pp. 209 - 214.

Sellier, P 2003, Protein nutrition for ruminants in european countries, in the light of

animal feeding regulations linkeed to bovine spongiform encephalothy. OIE Revue

Scientifique et Technique, vol. 22 no. 1 pp. 259-269.

Shen, Y, Fan, MZ, Ajakaiye, A & Archbold, T 2002, Use of the regression analysis

technique to determine the true phosphorus digestibility and the endogenous

phosphorus output associated with corn in growing pigs. The Journal of Nutrition, vol.

132 no. 6 pp.1199 - 1206.

Spencer, JD, Allee, GL & Sauber, TE 2000, Phosphorus bioavailability and

digestibility of normal and genetically modified low-phytate corn for pigs. Journal of

Animal Science, vol. 78 pp. 675 - 681.

Stein, HH, Kadzere, CT, Kim, SW & Miller, PS 2008, Influence of dietary phophorus

concentration on the digestibility of phosphorus in monocalcium phosphate by

growing pigs. Journal of Animal Science, vol. 86 pp. 1861 - 1867.

Thermo-Galactic 2003, PLSplus IQ User's Guide. ThermoGalactic, Salem, USA

Toor, GS, Peak, JD & Sims, JT 2005, Phosphorus Speciation in Broiler Litter and

Turky Manure Produced from Modified Diets. Journal of Environmental Quality, vol.

34 no. 2 pp. 687 - 697.

129

Traylor, SL, Cromwell, GL & Lindemann, MD 2005, Bioavailability of phosphorus in

meat and bone meal for swine. Journal of Animal Science. vol. 83 no. 5 pp. 1054-

1061.

Turner, BL 2004, Optimizing phosphorus characterization in animal manures by

solution phosphorus-31 nuclear magnetic resonance spectroscopy. Journal of

Environmental Quality, vol. 33 no.2 pp. 757-766.

Turner, BL & Leytem, AB 2004, Phosphorus Compounds in Sequential Extracts of

Animal Manures: Chemical Speciation and a Novel Fractionation Procedure.

Environmental Science and Technology, vol. 38 no.22 pp. 6101-6108.

Turner, BL & Richardson, AE 2004, Identification of scyllo-inositol phosphates in soil

by solution phosphorus-31 nuclear magnetic resonance spectroscopy. Soil Science

Society of America Journal, vol. 68 no. 6 pp. 802 -808.

Van Barneveld, RJ, Nuttall, JD, Flinn, PC & Osbourne, BG 1999, Near infrared

reflectance measurement fo the digestible energy content of cereals for growing pigs.

Journal of Near Infrared Spectroscopy, vol. 7 no. 1 pp. 1-7.

Van Kempen, T 2001, Infrared technology in animal production. World's Poultry

Science Journal, vol. 57 pp. 29 - 48.

Zyla, K, Ledoux, DR & Veum, TL 1995, Complete enzymatic dephosphorylation of

corn-soybean meal feed under simulated intestinal conditions of the turkey. Journal

of Agricultural and Food Chemistry, vol. 43 pp. 288 - 294.

130

Appendix A Percentage of Ca, K, Mg, Na, P and S in the Ingredient used (and an alternative

MBM) in the Pig Diets. Triplicate analysis of samples (C-1, C-2, and C-3)

Ingredient Element C-1 C-2 C-3 Mean Standard Deviation

Ca %w/w 24.97 21.04 24.63 24.5 0.5

K %w/w -0.003 0.013 0.043 0.02 0.02

Mg %w/w 1.306 1.282 1.267 1.29 0.02

Na %w/w 0.012 0.014 0.012 0.013 0.001

P %w/w 18.94 17.95 17.96 18.3 0.6

14257

REDIPHOS

DCP

S %w/w 0.871 0.860 0.867 0.866 0.005

Ca %w/w 29.08 28.35 28.88 28.8 0.4

K %w/w 0.0154 0.030 0.026 0.024 0.008

Mg %w/w 0.183 0.198 0.181 0.187 0.009

Na %w/w 0.026 0.025 0.025 0.025 0.001

P %w/w 19.00 18.28 18.89 18.7 0.4

14258

HONGHE

DCP

S %w/w 0.2673 0.2599 0.2647 0.264 0.004

Ca %w/w 5.58 6.16 5.98 5.9 0.3

K %w/w 0.784 0.798 0.807 0.80 0.01

Mg %w/w 0.166 0.199 0.196 0.19 0.02

Na %w/w 0.589 0.598 0.603 0.596 0.007

P %w/w 2.98 3.24 3.31 3.2 0.2

14259

QAF

MBM

S %w/w 0.540 0.556 0.548 0.548 0.008

131

Ingredient Element C-1 C-2 C-3 Mean Standard Deviation

Ca %w/w 37.5 38.8 39.6 38 1

K %w/w 0.087 0.091 0.098 0.092 0.005

Mg %w/w 0.305 0.315 0.258 0.29 0.03

Na %w/w 0.066 0.063 0.065 0.065 0.002

P %w/w 16.84 17.37 17.34 17.2 0.3

14260

PALPHOS

S %w/w 0.014 0.011 0.006 0.010 0.004

Ca %w/w 16.84 16.95 16.47 16.8 0.3

K %w/w 0.158 0.159 0.171 0.163 0.007

Mg %w/w 1.188 1.199 1.187 1.191 0.007

Na %w/w 0.188 0.191 0.183 0.187 0.004

P %w/w 20.76 21.38 21.22 21.1 0.3

14261

KYNOFOS

21

MDCP

S %w/w 0.743 0.740 0.732 0.738 0.006

Ca %w/w 23.97 24.23 24.81 24.3 0.4

K %w/w 1.013 1.043 0.981 1.01 0.03

Mg %w/w 0.056 0.064 0.063 0.061 0.004

Na %w/w 0.015 0.014 0.015 0.015 0.001

P %w/w 18.29 18.44 18.27 18.34 0.09

14262

BOLIFOR

DCP-S

S %w/w 0.028 0.037 0.024 0.030 0.007

132

Ingredient Element C-1 C-2 C-3 Mean Standard Deviation

Ca %w/w 0.059 0.061 0.056 0.058 0.003

K %w/w 0.131 0.138 0.134 0.135 0.003

Mg %w/w 0.009 0.010 0.009 0.009 0.001

Na %w/w 0.328 0.338 0.351 0.34 0.01

P %w/w 0.207 0.198 0.212 0.206 0.007

14263

BLOOD

MEAL

S %w/w 0.622 0.586 0.656 0.62 0.04

Ca %w/w 0.005 0.006 0.007 0.006 0.001

K %w/w 0.364 0.361 0.374 0.366 0.006

Mg %w/w 0.106 0.098 0.103 0.102 0.004

Na %w/w 0.001 0.000 -0.001 0.000 0.001

P %w/w 0.286 0.264 0.270 0.27 0.01

14264

CORN

S %w/w 0.112 0.105 0.104 0.107 0.004

Ca %w/w 21.80 20.90 22.61 21.8 0.9

K %w/w 0.535 0.510 0.516 0.52 0.01

Mg %w/w 3.183 3.042 3.048 3.09 0.08

Na %w/w 0.141 0.135 0.134 0.137 0.004

P %w/w 0.012 0.014 0.010 0.012 0.002

14265

LIMESTONE

S %w/w 0.010 0.010 0.010 0.010 0.000

133

Ingredient Element C-1 C-2 C-3 Mean Standard Deviation

Ca %w/w 6.08 6.41 6.64 6.4 0.3

K %w/w 0.360 0.354 0.328 0.35 0.02

Mg %w/w 0.150 0.147 0.150 0.149 0.002

Na %w/w 0.424 0.416 0.424 0.421 0.005

P %w/w 2.904 3.005 3.033 2.98 0.07

14266

MBM

S %w/w 0.633 0.606 0.636 0.63 0.02

Ca %w/w 0.406 0.451 0.396 0.42 0.03

K %w/w 2.55 2.76 2.86 2.7 0.2

Mg %w/w 0.318 0.323 0.300 0.31 0.01

Na %w/w 0.002 0.001 0.002 0.002 0.000

P %w/w 0.711 0.690 0.673 0.69 0.02

14437

SOYBEAN

MEAL

S %w/w 0.417 0.418 0.419 0.418 0.001

Ca %w/w 9.16 8.92 9.49 9.2 0.3

K %w/w 0.372 0.348 0.399 0.37 0.03

Mg %w/w 0.276 0.263 0.264 0.268 0.007

Na %w/w 0.043 0.040 0.043 0.042 0.001

P %w/w 0.038 0.044 0.043 0.042 0.003

14438

VITAMIN

MINERAL

PREMIX

S %w/w 1.531 1.392 1.456 1.46 0.07

134

Appendix B Example of printout and calculation from ‘FEEDMANIA DIET’ program FEEDMANIA DIET Created: Thu 9-Sep-04 10:06 am

2600 Phosphorus – Ray King Basal diet for P availability study – 0.55gAlys/MJDE INGREDIENTS INCLUDED Ingredient Amount Price Total Code % $/t Price $ DLMETH RES 0.058 1.160 kg 4280.00 4.96 LYSINE RES 0.160 3.203 kg 4200.00 13.45 THREO RES 0.031 0.624 kg 7400.00 4.62 MAIZE 7.5 70.284 1405.686 kg 0.00 0.00 STARCH 7.000 140.000 kg 0.00 0.00 SNFLWR OIL 2.000 40.000 kg 0.00 0.00 LIMES 2.066 41.326 kg 108.00 4.46 SALT 0.200 4.000 kg 292.00 1.17 VITA-GROW 0.200 4.000 kg 4012.00 16.05 BLD MEAL85 3.000 60.000 kg 650.00 39.00 SOY 48 15.000 300.000 kg 410.00 123.00 100.0 2000.000 kg Total 103.36 206.72 NUTRIENT COMPOSITION Nutrients Amount Code Name DE D.E. PIGS 14.58 MJ/kg PROT CRUDE PROTEIN 15.23 % LYS LYSINE (AVAILABLE) 0.802 % ISOL ISOLEUCINE 0.517 % LEU LEUCINE 1.517 % METHI METHIONINE 0.271 % MTHCY METHIONINE PLUS CYSTINE 0.520 % PHTYR PHENYLALANINE + TYROSINE 1.310 % THRE THREONINE 0.587 % TRYPT TRYPTOPHAN 0.186 % CA CALCIUM 0.800 % PAV PHOSPHORUS (AVAILABLE) 0.089 % FIBRE CRUDE FIBRE 2.657 % FEEDMANIA version 7.30.48 was created in AUSTRALIA by MANIA SOFTWARE Pty. Ltd. (c) 1984..93 Licensed User: VIAS .

135

Appendix C

Weight of individual feed ingredients (kg) as determined by ‘feedmania’.

Diet Number 1 2 3 4 5 6 7

Ingredient Basal Palphos

Rediphos

DCP

Honge

DCP

Bolifor

DCP Kynofos

QAF

MBM

Corn 214.614 107.307 107.307 107.307 107.307 107.307 107.307

Soybean meal 37.50 18.75 18.75 18.75 18.75 18.75 18.75

Blood meal 9.00 4.50 4.50 4.50 4.50 4.50 4.50

DL methionine 0.174 0.087 0.087 0.087 0.087 0.087 0.087

L. lysine 0.546 0.273 0.273 0.273 0.273 0.273 0.273

Mineral Vitamins 0.600 0.300 0.300 0.300 0.300 0.300 0.300

Salt 0.600 0.300 0.300 0.300 0.300 0.300 0.300

Starch 21 11.649 10.677 10.500 10.6635 10.221 4.6365

Sunflower Oil 6.0 3.0 3.0 3.4 3.0 3.0 3.0

Limestone 9.966 2.349 3.411 3.183 3.429 4.0545 2.802

Palphos 1.485

Rediphos DCP 1.395

Honge DCP 1.362

Bolifor DCP 1.3905

Kynofos 1.2075

QAF MBM 8.0445

Total Weight (kg) 300 150 150 150 150 150 150

136

Appendix D

Mineral results in the pig diets in mg/kg with mean and standard deviation of triplicate

analysis

Diet No. Diet 1 Diet 2 Diet 3 Diet 4 Diet 5 Diet 6 Diet 7

Diet type Basal PalphosRediphos

DCP

Honge

DCP

Bolifor

DCP Kynofor

QAF

MBM

mg/kg 7364 7479 7421 7116 7985 7555 8854 Calcium

SD 259 243 375 278 91 318 346

mg/kg 5387 5205 5206 5304 5423 5322 6799 Potassium

SD 203 160 234 49 25 122 116

mg/kg 2262 1748 2043 1869 2035 1947 1883 Magnesium

SD 97 66 63 60 28 69 48

mg/kg 793 807 678 855 954 995 1256 Sodium

SD 32 37 21 18 24 52 39

mg/kg 3127 4776 4591 4758 4827 4823 5304 Phosphorus

SD 88 131 139 129 169 154 100

mg/kg 1679 1587 1649 1640 1701 1695 2168 Sulphur

SD 49 40 50 51 21 61 65

137

Appendix E

Weights of Pigs in Kg on the date shown in 2004

Pig

ID

27-O

ct

5-N

ov

10-N

ov

19-N

ov

24-N

ov

3-D

ec

8-D

ec

17-D

ec

10/7 39.7 47.2 49.2 52.4 54.1 64.1 65.4 70.8

11/4 42.3 50.2 52.7 56.5 60.6 66.1 69.5 73.8

11/5 42.1 47.9 49.5 55.4 55.6 64.1 68.9 74.9

21/8 41.5 48.0 49.4 53.6 57.3 63.5 66.4 71.5

22/10 40.2 44.7 48.4 53.3 56.5 63.4 64.4 70.0

22/7 41.0 47.1 49.6 52.7 52.4 61.0 63.6 69.6

7/9 43.8 47.7 53.1 57.6 60.4 67.0 69.5 75.5

8/7 42.1 50.0 53.0 58.0 62.4 68.4 71.9 77.0

138

Appendix F

Comparison of the in vitro tests for P digestibility in the six P sources

Phosphorus Sources

In vitro test

Red

ipho

s D

CP

Hon

ghe

DC

P

Kyn

opho

s

MD

CP

Bol

ifor

DC

P

Pal

phos

QA

F

MB

M

(Mean and standard deviation in vitro P digestibility %)

1.08 0.25 66.2 0.65 0.23 4.0 CaCl2

0.02 0.02 0.5 0.04 0.03 0.1

4.9 0.81 43 2.33 0 4.9 Na HCO3 ½ hr

0.6 0.04 1 0.04 0.05 0.2

3.43 0.63 41.2 1.48 0.1 6.3 Na HCO3 16 hr

0.05 0.03 0.4 0.09 0.3 0.2

5.29 1.04 47.3 2.45 0 7.3 NH4 acetate

0.07 0.04 0.5 0.05 0.03 0.3

7.8 2.00 84 3.2 0.01 8.2 Water solubility

0.4 0.06 2 0.3 0 0.1

98 97 98 96 10.1 87.2 2% citric acid

1 2 3 2 0.2 0.4

9.8 3.54 42.8 7.7 2.0 18.1 HCl/ Bicarb 1.0 g

0.2 0.08 0.5 0.2 0.5 0.5

13.2 5.1 45.00 11.4 3.8 18.6 HCl/ Bicarb 0.75 g

0.3 0.2 0.03 0.1 0.2 0.5

18.0 8.58 47.1 17.0 5.66 28 HCl/ Bicarb 0.50 g

0.3 0.03 0.2 0.5 0.04 1

25.1 18.47 52 29.1 10.5 38 HCl/ Bicarb 0.25 g

0.6 0.04 1 0.2 0.6 3

In vivo P digestibility (%) 48.7 72.3 71.1 73.2 45.9 85.1

139

Comparison of the in vitro tests for P digestibility in the other P sources

Result in vitro P digestibility % of other Phosphorus Sources (single analysis)

In vitro test A

FP D

CP

Bio

fos

DC

P

Bio

phos

fish

mea

l

Mon

ofos

MA

P*

Mul

tifos

TC

P

Qua

ntum

DC

P

Qua

ntum

MD

CP

Red

ox D

CP

Red

ox M

DC

P

CaCl2 12.4 62.9 0 69.9 0 0 64.7 0 43.5

Na HCO3 ½ hr 7.9 38.6 0 82.2 0 0 41.4 4.4 27.2

Na HCO3 16 hr 5.5 39.1 0 77.1 0 0 45.0 3.3 25.3

NH4 acetate 7.9 41.6 0 78.8 1.1 0 41.4 3.8 29.6

Water solubility 14.9 80.1 0 89.4 0 0 83.9 4.4 50.6

2% citric acid 99.2 93.8 79.5 96.6 84.0 87.3 94.0 90.6 94.6

HCl/ Bicarb 1.0 g 11.4 34.0 4.1 74.5 3.4 3.3 37.9 8.7 26.8

HCl/ Bicarb .75 g 17.2 39.6 9.5 81.5 8.3 4.0 40.6 12.2 29.8

HCl/ Bicarb .50 g 19.0 43.0 15.5 80.1 11.2 6.2 45.1 16.2 34.5

HCl/ Bicarb .25 g 31.0 46.7 33.7 81.1 15.3 12.0 48.7 26.9 44.1

* MAP Mono-ammonium phosphate

140

Appendix G

MIR Spectrum

7800.0 7000 6000 5000 4000 3000 2000 1500 1000 450.00.45

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.61

cm-1

A

Kynophos MDCP and Bolifor DCP (Blue)

7800.0 7000 6000 5000 4000 3000 2000 1500 1000 450.00.44

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.78

cm-1

A

Rediphos DCP and Honge DCP(Blue)

141

7800.0 7000 6000 5000 4000 3000 2000 1500 1000 450.00.60

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2.0

2.1

2.2

2.3

2.4

2.5

2.60

cm-1

A

QAF - MBM and Palphos TCP(Blue)

7800.0 7000 6000 5000 4000 3000 2000 1500 1000 450.00.45

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0

3.11

cm-1

A

Biophos (fish meal + TCP) and Ridley DCP(Blue)

142

7800.0 7000 6000 5000 4000 3000 2000 1500 1000 450.00.48

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.82.88

cm-1

A

Biofos MDCP and Quantum MDCP(Blue)

143

Appendix H

MIR factor estimate curves for Digestibility

MIR factor estimate curves for Digestibility with 1st Derivative

144

MIR factor estimate curves for Total Phosphorus

MIR factor estimate curves for Total Phosphorus with 1st Derivative

145

MIR factor estimate curves for Total Calcium

MIR factor estimate curves for Total Calcium with 1st Derivative

146

Appendix I

SS-MSA-NMR

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Blood meal

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Corn

147

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Soybean meal

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Rediphos DCP

148

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Honge DCP

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

QAF – MBM

149

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Palphos TCP

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Kynofos MDCP

150

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Bolifor DCP

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – Basal Diet

151

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – Rediphos DCP

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – Honge DCP

152

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – QAF-MBM

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – Palphos TCP

153

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – Kynofos MDCP

-150-150-125-125-100-100-75-75-50-50-25-2500252550507575100100125125150150

Faeces – Bolifor DCP dihydrate